Re: [PATCH v5 4/5] vhost: add vmstate for inflight region with inner buffer

Stefan Hajnoczi <[email protected]> Mon, 12 Jan 2026 13:22:59 -0500
Newsgroups	dev.linux.lists.virtio-fs,org.nongnu.qemu-devel
Message-ID	<20260112182259.GF462084@fedora>
On Mon, Jan 12, 2026 at 04:45:02PM +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 <[email protected]>
> ---
>  hw/virtio/vhost.c         | 42 +++++++++++++++++++++++++++++++++++++++
>  include/hw/virtio/vhost.h |  6 ++++++
>  2 files changed, 48 insertions(+)
> 
> diff --git a/hw/virtio/vhost.c b/hw/virtio/vhost.c
> index c46203eb9c..9a746c9861 100644
> --- a/hw/virtio/vhost.c
> +++ b/hw/virtio/vhost.c
> @@ -2028,6 +2028,48 @@ 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()
> +    }
> +};
> +
> +const VMStateDescription vmstate_vhost_inflight_region = {
> +    .name = "vhost-inflight-region",
> +    .fields = (const VMStateField[]) {
> +        VMSTATE_UINT64(size, struct vhost_inflight),
> +        VMSTATE_UINT16(queue_size, struct vhost_inflight),
> +        VMSTATE_END_OF_LIST()
> +    },
> +    .subsections = (const VMStateDescription * const []) {
> +        &vmstate_vhost_inflight_region_buffer,
> +        NULL
> +    }
> +};

The subsection trick is neat - it allows the size to be loaded first and
then the memfd is allocated. However, it introduces a weird case: if the
source QEMU does not send the subsection, then the destination QEMU
loads successfully but with inflight partially uninitialized.

It's not obvious to me that the destination QEMU will fail in a safe way
when this happens. The source QEMU must not be able to trigger undefined
behavior. Can you add an explicit check somewhere to fail when this
required subsection is missing?

Thanks,
Stefan
signature.asc (application/pgp-signature, 488 B)