> Since the beginning of this thread I've tried to respond you and provide you
> all information I have, but all I heard from you is "show me", "I don't see
> it"
> and "I want it implemented my way, period".
>
> That's by far not productive.
I event sent you a working patch!
Anyways - I also g
On Mon, 17 Dec 2012 12:26:34 +
Dietmar Maurer wrote:
> > > No, there is no need to wait (see my patch).
> >
> > It has problems, replied to it.
>
> Please can you post a test case which show those problems? I simply can't
> reproduce any problem you mention.
Which problems are you referri
> > No, there is no need to wait (see my patch).
>
> It has problems, replied to it.
Please can you post a test case which show those problems? I simply can't
reproduce any problem you mention.
On Mon, 17 Dec 2012 12:17:28 +
Dietmar Maurer wrote:
> > Let me try once more to explain our options and why I think this is series
> > is
> > our best choice.
> >
> > Basically, we have try options:
> >
> > 1. Add stats to query-balloon
> >
> > This breaks existing clients, because query
> Let me try once more to explain our options and why I think this is series is
> our best choice.
>
> Basically, we have try options:
>
> 1. Add stats to query-balloon
>
> This breaks existing clients, because query-balloon is a synchronous
> command that returns immediately. If we add stats to
On Sat, 15 Dec 2012 07:19:13 +
Dietmar Maurer wrote:
> > - drop qmp & hmp interface of old stats code
>
> I think the old interface is not that bad (the new one is clumsy, because
> we need 6 qmp call instead of one to get all stats). Can't
> we try to make it functional and keep it?
No, b
> - drop qmp & hmp interface of old stats code
I think the old interface is not that bad (the new one is clumsy, because
we need 6 qmp call instead of one to get all stats). Can't
we try to make it functional and keep it?
Seriously, I think 'interval' should be a property we can pass at qemu
sta
This new try to re-enable the virtio-balloon driver stats uses QOM properties
via a polling mechanism as suggested by Anthony here:
http://lists.gnu.org/archive/html/qemu-devel/2012-02/msg02390.html
o v2
- drop qmp & hmp interface of old stats code
- allow polling callback to run even if the