Re: [Qemu-devel] [PATCHv2] virtio: verify that all outstanding buffers are flushed

[Michael S. Tsirkin]

On Thu, Dec 13, 2012 at 11:48:17AM +0100, Kevin Wolf wrote:
> Am 12.12.2012 17:37, schrieb Michael S. Tsirkin:
> > On Wed, Dec 12, 2012 at 04:52:53PM +0100, Paolo Bonzini wrote:
> >> Il 12/12/2012 16:25, Michael S. Tsirkin ha scritto:
> >>> On Wed, Dec 12, 2012 at 04:01:27PM +0100, Paolo Bonzini wrote:
> >>>> Il 12/12/2012 15:47, Michael S. Tsirkin ha scritto:
> >>>>> As step 1, I think we should just complete all outstanding
> >>>>> requests when VM stops.
> >>>>>
> >>>>> Yes it means you can't do the retry hack after migration
> >>>>> but this is hardly common scenario.
> >>>>
> >>>> I disagree...
> >>>
> >>> Disagree with what? You are saying it's common?
> >>
> >> It's not common, but you cannot block migration because you have an I/O
> >> error.  Solving the error may involve migrating the guests away from
> >> that host.
> >>
> >> Paolo
> > 
> > No, you should complete with error.
> 
> Just to confirm that this isn't possible: rerror/werror=stop is a
> supported feature, and I do get bug reports when it breaks (including
> migration while the VM is stopped for an I/O error).
> 
> Makes it common enough that breaking it is definitely not an option, right?
> 
> Kevin

Nod. So we need to fix inuse which is currently broken.
We still can have an assert but it needs to be done
more carefully, counting outstanding requests and
checking it matches inuse.

-- 
MST