On 2018/04/26 5:24, Jesper Dangaard Brouer wrote:
> On Tue, 24 Apr 2018 23:39:20 +0900
> Toshiaki Makita <toshiaki.maki...@gmail.com> wrote:
>
>> +static int veth_xdp_xmit(struct net_device *dev, struct xdp_frame *frame)
>> +{
>> + struct veth_priv *rcv_priv, *priv = netdev_priv(dev);
>> + int headroom = frame->data - (void *)frame;
>> + struct net_device *rcv;
>> + int err = 0;
>> +
>> + rcv = rcu_dereference(priv->peer);
>> + if (unlikely(!rcv))
>> + return -ENXIO;
>> +
>> + rcv_priv = netdev_priv(rcv);
>> + /* xdp_ring is initialized on receive side? */
>> + if (rcu_access_pointer(rcv_priv->xdp_prog)) {
>> + err = xdp_ok_fwd_dev(rcv, frame->len);
>> + if (unlikely(err))
>> + return err;
>> +
>> + err = veth_xdp_enqueue(rcv_priv, veth_xdp_to_ptr(frame));
>> + } else {
>> + struct sk_buff *skb;
>> +
>> + skb = veth_build_skb(frame, headroom, frame->len, 0);
>> + if (unlikely(!skb))
>> + return -ENOMEM;
>> +
>> + /* Get page ref in case skb is dropped in netif_rx.
>> + * The caller is responsible for freeing the page on error.
>> + */
>> + get_page(virt_to_page(frame->data));
>
> I'm not sure you can make this assumption, that xdp_frames coming from
> another device driver uses a refcnt based memory model. But maybe I'm
> confused, as this looks like an SKB receive path, but in the
> ndo_xdp_xmit().
I find this path similar to cpumap, which creates skb from redirected
xdp frame. Once it is converted to skb, skb head is freed by
page_frag_free, so anyway I needed to get the refcount here regardless
of memory model.
>
>
>> + if (unlikely(veth_forward_skb(rcv, skb, false) !=
>> NET_RX_SUCCESS))
>> + return -ENXIO;
>> +
>> + /* Put page ref on success */
>> + page_frag_free(frame->data);
>> + }
>> +
>> + return err;
>> +}
>
>
--
Toshiaki Makita
