vhost-user: Fix the virtio features negotiation flaw

This patch aims to fix unexpected negotiation features for
vhost-user netdev interface.

When openvswitch reconnect Qemu after an unexpected disconnection
and Qemu therefore start the vhost_dev, acked_features field in
vhost_dev is initialized with value fetched from acked_features
field in NetVhostUserState, which should be up-to-date at that
moment but Qemu could not make it actually during the time window
of virtio features negotiation.

So we save the acked_features right after being configured by
guest virtio driver so it can be used to restore acked_features
field in vhost_dev correctly.

Signed-off-by: Hyman Huang(黄勇) <huangy81@chinatelecom.cn>
Signed-off-by: Guoyi Tu <tugy@chinatelecom.cn>
Signed-off-by: Liuxiangdong <liuxiangdong5@huawei.com>
Message-Id: <b9f8cf5561a79ea65ea38960e5a5e6d3707eef0a.1671627406.git.huangy81@chinatelecom.cn>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
This commit is contained in:
Hyman Huang(黄勇) 2022-12-21 21:06:40 +08:00 committed by Michael S. Tsirkin
parent bebcac052a
commit c9bdc449f9
4 changed files with 22 additions and 0 deletions

View file

@ -984,6 +984,12 @@ static void virtio_net_set_features(VirtIODevice *vdev, uint64_t features)
continue;
}
vhost_net_ack_features(get_vhost_net(nc->peer), features);
/*
* keep acked_features in NetVhostUserState up-to-date so it
* can't miss any features configured by guest virtio driver.
*/
vhost_net_save_acked_features(nc->peer);
}
if (virtio_has_feature(features, VIRTIO_NET_F_CTRL_VLAN)) {