On Thu,  7 May 2020 15:40:39 +0300 Denis Kirjanov wrote:
> The patch adds a basic XDP processing to xen-netfront driver.
> 
> We ran an XDP program for an RX response received from netback
> driver. Also we request xen-netback to adjust data offset for
> bpf_xdp_adjust_head() header space for custom headers.
> 
> synchronization between frontend and backend parts is done
> by using xenbus state switching:
> Reconfiguring -> Reconfigured- > Connected
> 
> UDP packets drop rate using xdp program is around 310 kpps
> using ./pktgen_sample04_many_flows.sh and 160 kpps without the patch.

Please provide a cover letter for the submission.

> @@ -167,6 +179,9 @@ struct netfront_rx_info {
>       struct xen_netif_extra_info extras[XEN_NETIF_EXTRA_TYPE_MAX - 1];
>  };
>  
> +static int xennet_xdp_xmit(struct net_device *dev, int n,
> +                        struct xdp_frame **frames, u32 flags);

Is it possible to put the function here and avoid the forward
declaration?

>  static void skb_entry_set_link(union skb_entry *list, unsigned short id)
>  {
>       list->link = id;
> @@ -265,8 +280,8 @@ static struct sk_buff *xennet_alloc_one_rx_buffer(struct 
> netfront_queue *queue)
>       if (unlikely(!skb))
>               return NULL;
>  
> -     page = alloc_page(GFP_ATOMIC | __GFP_NOWARN);
> -     if (!page) {
> +     page = page_pool_dev_alloc_pages(queue->page_pool);
> +     if (unlikely(!page)) {
>               kfree_skb(skb);
>               return NULL;
>       }
> @@ -778,6 +793,53 @@ static int xennet_get_extras(struct netfront_queue 
> *queue,
>       return err;
>  }
>  
> +u32 xennet_run_xdp(struct netfront_queue *queue, struct page *pdata,

static

> +                struct xen_netif_rx_response *rx, struct bpf_prog *prog,
> +                struct xdp_buff *xdp)
> +{
> +     struct xdp_frame *xdpf;
> +     u32 len = rx->status;
> +     u32 act = XDP_PASS;
> +     int err;
> +
> +     xdp->data_hard_start = page_address(pdata);
> +     xdp->data = xdp->data_hard_start + XDP_PACKET_HEADROOM;
> +     xdp_set_data_meta_invalid(xdp);
> +     xdp->data_end = xdp->data + len;
> +     xdp->rxq = &queue->xdp_rxq;
> +     xdp->handle = 0;

No need to clear this.

> +     act = bpf_prog_run_xdp(prog, xdp);
> +     switch (act) {
> +     case XDP_TX:
> +             get_page(pdata);
> +             xdpf = convert_to_xdp_frame(xdp);
> +             err = xennet_xdp_xmit(queue->info->netdev, 1, &xdpf, 0);
> +             if (unlikely(err < 0))
> +                     trace_xdp_exception(queue->info->netdev, prog, act);
> +             break;
> +     case XDP_REDIRECT:
> +             get_page(pdata);
> +             err = xdp_do_redirect(queue->info->netdev, xdp, prog);
> +             if (unlikely(err))
> +                     trace_xdp_exception(queue->info->netdev, prog, act);
> +             xdp_do_flush();

Can you call flush after the RX loop? (xennet_poll()?) This is supposed
to batch all redirections in the NAPI cycle.

> +             break;
> +     case XDP_PASS:
> +     case XDP_DROP:
> +             break;
> +
> +     case XDP_ABORTED:
> +             trace_xdp_exception(queue->info->netdev, prog, act);
> +             break;
> +
> +     default:
> +             bpf_warn_invalid_xdp_action(act);
> +     }
> +
> +     return act;
> +}
> +
>  static int xennet_get_responses(struct netfront_queue *queue,
>                               struct netfront_rx_info *rinfo, RING_IDX rp,
>                               struct sk_buff_head *list)
> @@ -792,6 +854,9 @@ static int xennet_get_responses(struct netfront_queue 
> *queue,
>       int slots = 1;
>       int err = 0;
>       unsigned long ret;
> +     struct bpf_prog *xdp_prog;
> +     struct xdp_buff xdp;
> +     u32 verdict;
>  
>       if (rx->flags & XEN_NETRXF_extra_info) {
>               err = xennet_get_extras(queue, extras, rp);
> @@ -827,9 +892,20 @@ static int xennet_get_responses(struct netfront_queue 
> *queue,
>  
>               gnttab_release_grant_reference(&queue->gref_rx_head, ref);
>  
> -             __skb_queue_tail(list, skb);
> -
> +             rcu_read_lock();
> +             xdp_prog = rcu_dereference(queue->xdp_prog);
> +             if (xdp_prog && !(rx->flags & XEN_NETRXF_more_data)) {

What is XEN_NETRXF_more_data?

If you have a mis-formatted frame it has to be dropped, nothing can
escape the XDP program and go to the stack.

> +                     /* currently only a single page contains data */
> +                     WARN_ON_ONCE(skb_shinfo(skb)->nr_frags != 1);
> +                     verdict = xennet_run_xdp(queue,
> +                                    
> skb_frag_page(&skb_shinfo(skb)->frags[0]),
> +                                    rx, xdp_prog, &xdp);
> +                     if (verdict != XDP_PASS)
> +                             err = -EINVAL;
> +             }
> +             rcu_read_unlock();
>  next:
> +             __skb_queue_tail(list, skb);
>               if (!(rx->flags & XEN_NETRXF_more_data))
>                       break;
>  

> +static int xennet_xdp_set(struct net_device *dev, struct bpf_prog *prog,
> +                     struct netlink_ext_ack *extack)
> +{
> +     struct netfront_info *np = netdev_priv(dev);
> +     struct bpf_prog *old_prog;
> +     unsigned int i, err;
> +     unsigned long int max_mtu = XEN_PAGE_SIZE - XDP_PACKET_HEADROOM;
> +
> +     if (dev->mtu > max_mtu) {
> +             netdev_warn(dev, "XDP requires MTU less than %lu\n", max_mtu);
> +             return -EINVAL;
> +     }
> +
> +     if (!np->netback_has_xdp_headroom)
> +             return 0;
> +
> +     old_prog = rtnl_dereference(np->queues[0].xdp_prog);
> +     if (!old_prog && !prog)

You don't have to check this, the core will not call drivers any more
if there is no program and request is to disable XDP.

> +             return 0;
> +
> +     if (prog)
> +             bpf_prog_add(prog, dev->real_num_tx_queues);
> +
> +     for (i = 0; i < dev->real_num_tx_queues; ++i)
> +             rcu_assign_pointer(np->queues[i].xdp_prog, prog);

I'm I reading this right the moment you assign the program the RX
function will start assuming there is a headroom, even if you haven't
reconfigured the backend, yet..

> +     if (old_prog)
> +             for (i = 0; i < dev->real_num_tx_queues; ++i)
> +                     bpf_prog_put(old_prog);
> +
> +     xenbus_switch_state(np->xbdev, XenbusStateReconfiguring);
> +
> +     err = talk_to_netback_xdp(np, prog ? NETBACK_XDP_HEADROOM_ENABLE:
> +                               NETBACK_XDP_HEADROOM_DISABLE);
> +     if (err)
> +             return err;
> +
> +     /* avoid race with XDP headroom adjustment */
> +     wait_event(module_wq,
> +                xenbus_read_driver_state(np->xbdev->otherend) ==
> +                XenbusStateReconfigured);
> +     np->netfront_xdp_enabled = true;
> +     xenbus_switch_state(np->xbdev, XenbusStateConnected);
> +
> +     return 0;
> +}
> +

> @@ -1754,6 +1993,51 @@ static void xennet_destroy_queues(struct netfront_info 
> *info)
>       info->queues = NULL;
>  }
>  
> +
> +
> +static int xennet_create_page_pool(struct netfront_queue *queue)

Please run checkpatch --strict on your patches.

Reply via email to