[libvirt] [PATCH] migration: remove direct migration dependency on version1 of driver

Nikolay Shirokovskiy nshirokovskiy at parallels.com
Mon Aug 31 08:42:13 UTC 2015



On 28.08.2015 19:04, Michal Privoznik wrote:
> On 28.08.2015 11:29, Nikolay Shirokovskiy wrote:
>>
>>
>> On 28.08.2015 08:54, Michal Privoznik wrote:
>>> On 27.08.2015 12:23, Nikolay Shirokovskiy wrote:
>>>> From: Nikolay Shirokovskiy <Nikolay Shirokovskiy nshirokovskiy at virtuozzo.com>
>>>>
>>>> Direct migration should work if *perform3 is present but *perform
>>>> is not. This is situation when driver migration is implemented
>>>> after new version of driver function is introduced. We should not
>>>> be forced to support old version too as its parameter space is
>>>> subspace of newer one.
>>>>
>>>> Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy at virtuozzo.com>
>>>> ---
>>>>  src/libvirt-domain.c |    3 ++-
>>>>  1 files changed, 2 insertions(+), 1 deletions(-)
>>>>
>>>> diff --git a/src/libvirt-domain.c b/src/libvirt-domain.c
>>>> index 6ab50ba..c89775b 100644
>>>> --- a/src/libvirt-domain.c
>>>> +++ b/src/libvirt-domain.c
>>>> @@ -3427,7 +3427,8 @@ virDomainMigrateDirect(virDomainPtr domain,
>>>>                       NULLSTR(xmlin), flags, NULLSTR(dname), NULLSTR(dconnuri),
>>>>                       NULLSTR(miguri), bandwidth);
>>>>  
>>>> -    if (!domain->conn->driver->domainMigratePerform) {
>>>> +    if (!domain->conn->driver->domainMigratePerform &&
>>>> +        !domain->conn->driver->domainMigratePerform3) {
>>>>          virReportUnsupportedError();
>>>>          return -1;
>>>>      }
>>>>
>>>
>>>
>>> Hm.. domainMigratePerform3 will be used iff connection driver has
>>> VIR_DRV_FEATURE_MIGRATION_V3 feature. But this check will require that
>>> regardless. What if we check the presence of implementation with respect
>>> to that?
>> I see you mean actual driver could be behind remote one and checking
>> for perform3 always gives true so we need to check for feature instead?
> 
> Sort of. domainMigratePerform3 is used only if the feature is present.
> But after your patch, the function implementation would be needed
> always, even if the feature is not present.
Why? You can implement version1 or version3 or both for this check to pass.
> 
> Michal
> 




More information about the libvir-list mailing list