Hi Jesper,
On Wed, Oct 3, 2018 at 7:04 AM Jesper Dangaard Brouer <bro...@redhat.com> wrote:
>
> On Tue, 25 Sep 2018 22:36:39 -0700
> Song Liu <liu.song....@gmail.com> wrote:
>
> > On Tue, Sep 25, 2018 at 7:26 AM Jesper Dangaard Brouer
> > <bro...@redhat.com> wrote:
> > >
> > > XDP can modify (and resize) the Ethernet header in the packet.
> > >
> > > There is a bug in generic-XDP, because skb->protocol and skb->pkt_type
> > > are setup before reaching (netif_receive_)generic_xdp.
> > >
> > > This bug was hit when XDP were popping VLAN headers (changing
> > > eth->h_proto), as skb->protocol still contains VLAN-indication
> > > (ETH_P_8021Q) causing invocation of skb_vlan_untag(skb), which corrupt
> > > the packet (basically popping the VLAN again).
> > >
> > > This patch catch if XDP changed eth header in such a way, that SKB
> > > fields needs to be updated.
> > >
> > > Fixes: d445516966dc ("net: xdp: support xdp generic on virtual devices")
> > > Signed-off-by: Jesper Dangaard Brouer <bro...@redhat.com>
> > > ---
> > > net/core/dev.c | 14 ++++++++++++++
> > > 1 file changed, 14 insertions(+)
> > >
> > > diff --git a/net/core/dev.c b/net/core/dev.c
> > > index ca78dc5a79a3..db6d89f536cb 100644
> > > --- a/net/core/dev.c
> > > +++ b/net/core/dev.c
> > > @@ -4258,6 +4258,9 @@ static u32 netif_receive_generic_xdp(struct sk_buff
> > > *skb,
> > > struct netdev_rx_queue *rxqueue;
> > > void *orig_data, *orig_data_end;
> > > u32 metalen, act = XDP_DROP;
> > > + __be16 orig_eth_type;
> > > + struct ethhdr *eth;
> > > + bool orig_bcast;
> > > int hlen, off;
> > > u32 mac_len;
> > >
> > > @@ -4298,6 +4301,9 @@ static u32 netif_receive_generic_xdp(struct sk_buff
> > > *skb,
> > > xdp->data_hard_start = skb->data - skb_headroom(skb);
> > > orig_data_end = xdp->data_end;
> > > orig_data = xdp->data;
> > > + eth = (struct ethhdr *)xdp->data;
> > > + orig_bcast = is_multicast_ether_addr_64bits(eth->h_dest);
> > > + orig_eth_type = eth->h_proto;
> > >
> > > rxqueue = netif_get_rxqueue(skb);
> > > xdp->rxq = &rxqueue->xdp_rxq;
> > > @@ -4321,6 +4327,14 @@ static u32 netif_receive_generic_xdp(struct
> > > sk_buff *skb,
> > >
> > > }
> > >
> > > + /* check if XDP changed eth hdr such SKB needs update */
> > > + eth = (struct ethhdr *)xdp->data;
> > > + if ((orig_eth_type != eth->h_proto) ||
> > > + (orig_bcast != is_multicast_ether_addr_64bits(eth->h_dest))) {
> >
> > Is the actions below always correct for the condition above? Do we need
> > to confirm the SKB is updated properly?
>
> I cannot find the issue that you are hinting to?
>
> If the BPF prog used bpf_xdp_adjust_head(), which the included selftest
> program does, then skb->data have been appropriately adjusted just
> above (with __skb_pull(skb, off) or __skb_push(skb, -off)), which makes
> the call to skb_reset_mac_header(skb) inside eth_type_trans() correct.
>
> I've double checked the code, and I cannot find anything wrong...
> please let me know if I missed something!?
>
>
> > > + __skb_push(skb, mac_len);
> > > + skb->protocol = eth_type_trans(skb, skb->dev);
>
> We could change mac_len to be ETH_HLEN, because inside eth_type_trans()
> the constant ETH_HLEN is used, that way we are 100% sure the
> skb_push/skb_pull are "paired". Will that be better for you?
Thanks for the explanation. Now I understand it better.
I think using ETH_HLEN is better. Both napi_frags_finish() and
__dev_forward_skb()
use similar pattern.
Song
