vhost: configure vDPA device after set vring call message
authorXiaolong Ye <xiaolong.ye@intel.com>
Mon, 14 Jan 2019 07:34:24 +0000 (15:34 +0800)
committerFerruh Yigit <ferruh.yigit@intel.com>
Fri, 18 Jan 2019 08:47:26 +0000 (09:47 +0100)
As qemu will only send VHOST_USER_SET_VRING_ENABLE message for guest
enabled vrings (only first queue pair will be enabled at initialized
stage), this will cause trouble for multiqueue case, vDPA's dev_conf
callback will get no chance be invoked. Decouple the dev_conf callback from
VHOST_USER_SET_VRING_ENABLE solves this issue.

Signed-off-by: Xiaolong Ye <xiaolong.ye@intel.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
lib/librte_vhost/vhost_user.c

index 1843e03..f2f3036 100644 (file)
@@ -2048,7 +2048,7 @@ skip_to_reply:
        vdpa_dev = rte_vdpa_get_device(did);
        if (vdpa_dev && virtio_is_ready(dev) &&
                        !(dev->flags & VIRTIO_DEV_VDPA_CONFIGURED) &&
-                       msg.request.master == VHOST_USER_SET_VRING_ENABLE) {
+                       msg.request.master == VHOST_USER_SET_VRING_CALL) {
                if (vdpa_dev->ops->dev_conf)
                        vdpa_dev->ops->dev_conf(dev->vid);
                dev->flags |= VIRTIO_DEV_VDPA_CONFIGURED;