On Tue, Mar 7, 2023 at 6:02 PM Michael S. Tsirkin <m...@redhat.com> wrote: > > On Tue, Mar 07, 2023 at 06:00:18PM +0100, Eugenio Pérez wrote: > > There is no reason to block it as it has nothing to do with the vrings. > > All the support of the feature comes via config space. > > > > Signed-off-by: Eugenio Pérez <epere...@redhat.com> > > Suggested-by: Alvaro Karsz <alvaro.ka...@solid-run.com> > > do we need to version this with machine type btw? >
No, as far as I know. If the destination qemu does not support VIRTIO_NET_F_SPEED_DUPLEX and it enables SVQ it will refuse to start. Thanks! > > --- > > net/vhost-vdpa.c | 3 ++- > > 1 file changed, 2 insertions(+), 1 deletion(-) > > > > diff --git a/net/vhost-vdpa.c b/net/vhost-vdpa.c > > index de5ed8ff22..8b25559320 100644 > > --- a/net/vhost-vdpa.c > > +++ b/net/vhost-vdpa.c > > @@ -99,7 +99,8 @@ static const uint64_t vdpa_svq_device_features = > > BIT_ULL(VIRTIO_F_ANY_LAYOUT) | > > BIT_ULL(VIRTIO_NET_F_CTRL_MAC_ADDR) | > > BIT_ULL(VIRTIO_NET_F_RSC_EXT) | > > - BIT_ULL(VIRTIO_NET_F_STANDBY); > > + BIT_ULL(VIRTIO_NET_F_STANDBY) | > > + BIT_ULL(VIRTIO_NET_F_SPEED_DUPLEX); > > > > #define VHOST_VDPA_NET_CVQ_ASID 1 > > > > -- > > 2.31.1 >
