On 12/06/2016 04:04 AM, Stefan Hajnoczi wrote:

>>
>> Prasanna Kumar Kalever (3):
>>   block/gluster: fix QMP to match debug option
>>   block/nfs: fix QMP to match debug option
>>   qemu-doc: update gluster protocol usage guide
>>
>>  block/gluster.c      | 38 ++++++++++++++++-----------------
>>  block/nfs.c          |  4 ++--
>>  qapi/block-core.json |  8 +++----
>>  qemu-doc.texi        | 59 
>> +++++++++++++++++++++++++++++++++++++++-------------
>>  qemu-options.hx      | 25 ++++++++++++++++++++--
>>  5 files changed, 93 insertions(+), 41 deletions(-)
> 
> BlockdevOptionsGluster.debug(-level) does not have "Added in 2.8" so I
> had to dig through git-blame(1) to verify that it was indeed added in
> the current release cycle.

Then that implies we should add yet one more patch that adds the
appropriate versioning information to all the gluster fields added for
2.8.  My reviewed-by was given on the assumption that debug was in 2.7
and that this was a break from 2.7 behavior, but that we already KNOW
we're breaking blockdev-add between 2.7 and 2.8; while your argument is
that there is no backwards incompatibility because it was not in 2.7 to
begin with.  I think both reasons are indeed acceptable, but it also
means that my reason was flawed because of the incomplete documentation.

> 
> In the future please make sure all QAPI changes are marked by version.

Indeed, and I try to flag it in my reviews as often as I notice it.

> If there tricky changes you can include a statement showing you are
> aware of QAPI backwards compatibility ("These new options were added in
> the 2.8 release cycle and can therefore still be changed without
> breaking backward compatibility").  This will make me confident that
> you've checked the QAPI changes.
> 
> Thanks, applied to my staging tree:
> https://github.com/stefanha/qemu/commits/staging
> 
> Stefan
> 

-- 
Eric Blake   eblake redhat com    +1-919-301-3266
Libvirt virtualization library http://libvirt.org

Attachment: signature.asc
Description: OpenPGP digital signature

Reply via email to