Re: [PATCH v6 4/5] vhost: add vmstate for inflight region with inner buffer

[Peter Xu <peterx@redhat.com>]

On Wed, Jan 14, 2026 at 04:38:17PM -0500, Stefan Hajnoczi wrote:
> On Wed, Jan 14, 2026 at 02:15:27PM -0500, Peter Xu wrote:
> > On Tue, Jan 13, 2026 at 02:58:17PM +0500, Alexandr Moshkov wrote:
> > > Prepare for future inflight region migration for vhost-user-blk.
> > > We need to migrate size, queue_size, and inner buffer.
> > > 
> > > So firstly it migrate size and queue_size fields, then allocate memory for buffer with
> > > migrated size, then migrate inner buffer itself.
> > > 
> > > Signed-off-by: Alexandr Moshkov <dtalexundeer@yandex-team.ru>
> > > ---
> > >  hw/virtio/vhost.c         | 57 +++++++++++++++++++++++++++++++++++++++
> > >  include/hw/virtio/vhost.h |  6 +++++
> > >  2 files changed, 63 insertions(+)
> > > 
> > > diff --git a/hw/virtio/vhost.c b/hw/virtio/vhost.c
> > > index c46203eb9c..f655c53b67 100644
> > > --- a/hw/virtio/vhost.c
> > > +++ b/hw/virtio/vhost.c
> > > @@ -2028,6 +2028,63 @@ const VMStateDescription vmstate_backend_transfer_vhost_inflight = {
> > >      }
> > >  };
> > >  
> > > +static int vhost_inflight_buffer_pre_load(void *opaque, Error **errp)
> > > +{
> > > +    info_report("vhost_inflight_region_buffer_pre_load");
> > > +    struct vhost_inflight *inflight = opaque;
> > > +
> > > +    int fd = -1;
> > > +    void *addr = qemu_memfd_alloc("vhost-inflight", inflight->size,
> > > +                                  F_SEAL_GROW | F_SEAL_SHRINK | F_SEAL_SEAL,
> > > +                                  &fd, errp);
> > > +    if (*errp) {
> > > +        return -ENOMEM;
> > > +    }
> > > +
> > > +    inflight->offset = 0;
> > > +    inflight->addr = addr;
> > > +    inflight->fd = fd;
> > > +
> > > +    return 0;
> > > +}
> > > +
> > > +const VMStateDescription vmstate_vhost_inflight_region_buffer = {
> > > +    .name = "vhost-inflight-region/buffer",
> > > +    .pre_load_errp = vhost_inflight_buffer_pre_load,
> > > +    .fields = (const VMStateField[]) {
> > > +        VMSTATE_VBUFFER_UINT64(addr, struct vhost_inflight, 0, NULL, size),
> > > +        VMSTATE_END_OF_LIST()
> > > +    }
> > > +};
> > > +
> > > +static int vhost_inflight_region_post_load(void *opaque,
> > > +                                           int version_id,
> > > +                                           Error **errp)
> > > +{
> > > +    struct vhost_inflight *inflight = opaque;
> > > +
> > > +    if (inflight->addr == NULL) {
> > 
> > IIUC this can never happen because pre_load() must trigger before
> > post_load(), and when reaching post_load() it means pre_load() must have
> > succeeded..
> > 
> > So, IIUC we can drop this post_load() completely (or assert addr in
> > pre_load instead).
> 
> I asked for this input validation check. If the migration stream is
> inconsistent (e.g. broken or malicious source QEMU), then the subsection
> might be missing but size could be non-zero. The destination QEMU should
> fail cleanly and not run into undefined behavior.

Ah I misread it as the one pairing with the pre_load().  It makes sense
indeed to have such post_load() in the parent VMSD.

Please ignore my comment, sorry for the noise.

-- 
Peter Xu