Re: [Qemu-devel] [PATCH] virtio-pci: Fix cross-version migration with older machines

[Cornelia Huck]

On Wed, 14 Dec 2016 15:16:13 +0200
"Michael S. Tsirkin" <address@hidden> wrote:

> On Wed, Dec 14, 2016 at 02:12:10PM +0100, Cornelia Huck wrote:
> > On Wed, 14 Dec 2016 13:52:37 +0100
> > Maxime Coquelin <address@hidden> wrote:
> > 
> > > This patch fixes a cross-version migration regression introduced
> > > by commit d1b4259f ("virtio-bus: Plug devices after features are
> > > negotiated").
> > > 
> > > The problem is encountered when host's vhost backend does not support
> > > VIRTIO_F_VERSION_1, and migration is initiated from a v2.7 or prior
> > > machine with virtio-pci modern capabilities enabled to a v2.8 machine.
> > 
> > Wait, machine versions or qemu versions?
> > 
> > > 
> > > In this case, modern capabilities get exposed to the guest by the source,
> > > whereas the target will detect version 1 is not supported so will only
> > > expose legacy capabilities.
> > > 
> > > The problem is fixed by introducing a new "x-modern-broken" property,
> > > which is set in v2.7 and prior compatibility modes. Doing this, v2.7
> > > machine keeps its broken behaviour (enabling modern while version is
> > > not supported), and newer machines will behave correctly.
> > > 
> > > Reported-by: Michael Roth <address@hidden>
> > > Suggested-by: Stefan Hajnoczi <address@hidden>
> > > Cc: Michael S. Tsirkin <address@hidden>
> > > Cc: Cornelia Huck <address@hidden>
> > > Cc: Marcel Apfelbaum <address@hidden>
> > > Cc: Dr. David Alan Gilbert <address@hidden>
> > > Signed-off-by: Maxime Coquelin <address@hidden>
> > > ---
> > > 
> > > I'm not sure about the property name, let me know if you have better 
> > > ideas.
> > > I didn't tested migration yet, but I wanted to share the patch while I 
> > > test it.
> > > I tested booting v2.8 and v2.7 machines with !VERSION_1 and get expected 
> > > result:
> > >  - v2.8: Virtio-pci probe succeed
> > >  - v2.7: Virtio-pci probe fails
> > > 
> > > Thanks,
> > > Maxime
> > > 
> > >  hw/virtio/virtio-pci.c | 4 +++-
> > >  hw/virtio/virtio-pci.h | 1 +
> > >  include/hw/compat.h    | 4 ++++
> > >  3 files changed, 8 insertions(+), 1 deletion(-)
> > > 
> > > diff --git a/hw/virtio/virtio-pci.c b/hw/virtio/virtio-pci.c
> > > index 521ba0b..93f6b54 100644
> > > --- a/hw/virtio/virtio-pci.c
> > > +++ b/hw/virtio/virtio-pci.c
> > > @@ -1580,7 +1580,8 @@ static void virtio_pci_device_plugged(DeviceState 
> > > *d, Error **errp)
> > >       * Virtio capabilities present without
> > >       * VIRTIO_F_VERSION_1 confuses guests
> > >       */
> > > -    if (!virtio_has_feature(vdev->host_features, VIRTIO_F_VERSION_1)) {
> > > +    if (!proxy->modern_broken &&
> > 
> > What you are de facto doing here is ignoring the features supported by
> > the backend. Call this ->ignore_backend_features or so?
> > 
> > > +            !virtio_has_feature(vdev->host_features, 
> > > VIRTIO_F_VERSION_1)) {
> > >          virtio_pci_disable_modern(proxy);
> > > 
> > >          if (!legacy) {
> > > @@ -1852,6 +1853,7 @@ static Property virtio_pci_properties[] = {
> > >                      VIRTIO_PCI_FLAG_DISABLE_PCIE_BIT, false),
> > >      DEFINE_PROP_BIT("page-per-vq", VirtIOPCIProxy, flags,
> > >                      VIRTIO_PCI_FLAG_PAGE_PER_VQ_BIT, false),
> > > +    DEFINE_PROP_BOOL("x-modern-broken", VirtIOPCIProxy, modern_broken, 
> > > false),
> > 
> > What about "x-ignore-backend-features"?
> 
> It's just the capability that ignores that backend is legacy.
> 
> x-cap-ignore-backend-legacy
> 
> ?

Sounds good. And has the benefit that nobody will be tempted to set
that manually :)