On Tue, Mar 28, 2017 at 05:29:34PM +0200, Kashyap Chamarthy wrote:
> On Tue, Mar 28, 2017 at 09:49:40AM -0500, Eric Blake wrote:
> > On 03/24/2017 07:34 AM, Kashyap Chamarthy wrote:
> > > While debugging some other issue, I happened to stumble across an old
> > > libvirt commit[*] that adds support for pivot (whether QEMU should
> > > switch to a target copy or not) operation as a result of issuing QMP
> > > 'block-job-cancel' to a 'drive-mirror' (in libvirt parlance, "block
> > > copy").
> > > 
> > > In the libvirt commit message[*] Eric Blake writes:
> > > 
> > >     "[...] There may be potential improvements to the snapshot code to
> > >     exploit block copy over multiple disks all at one point in time.
> > >     And, if 'block-job-cancel' were made part of 'transaction', you
> > >     could copy multiple disks at the same point in time without pausing
> > >     the domain. [...]"
> > > 
> > > I realize that 'block-job-cancel' is currently not part of the
> > > @TransactionAction.  Is it worthwhile to do so?
> > 
> > Yes, it is still worthwhile to make that improvement, although it is
> > now 2.10 material.
> 
> Thanks for looking, didn't imply that it's for 2.9 (which is in Release
> Candidate phase) -- I realize most people are busy with it.  Further
> discussion could be had when 2.10 tree opens up.

Thanks for the discussion.  I have added this task to the block layer
todo list:

http://wiki.qemu-project.org/ToDo/Block#Atomic_block-job-cancel