Jesper Dangaard Brouer wrote:
> This BPF-helper bpf_check_mtu() works for both XDP and TC-BPF programs.
>
> The SKB object is complex and the skb->len value (accessible from
> BPF-prog) also include the length of any extra GRO/GSO segments, but
> without taking into account that these GRO/GSO segments get added
> transport (L4) and network (L3) headers before being transmitted. Thus,
> this BPF-helper is created such that the BPF-programmer don't need to
> handle these details in the BPF-prog.
>
> The API is designed to help the BPF-programmer, that want to do packet
> context size changes, which involves other helpers. These other helpers
> usually does a delta size adjustment. This helper also support a delta
> size (len_diff), which allow BPF-programmer to reuse arguments needed by
> these other helpers, and perform the MTU check prior to doing any actual
> size adjustment of the packet context.
>
> It is on purpose, that we allow the len adjustment to become a negative
> result, that will pass the MTU check. This might seem weird, but it's not
> this helpers responsibility to "catch" wrong len_diff adjustments. Other
> helpers will take care of these checks, if BPF-programmer chooses to do
> actual size adjustment.
>
> V13:
> - Enforce flag BPF_MTU_CHK_SEGS cannot use len_diff.
>
> V12:
> - Simplify segment check that calls skb_gso_validate_network_len.
> - Helpers should return long
>
> V9:
> - Use dev->hard_header_len (instead of ETH_HLEN)
> - Annotate with unlikely req from Daniel
> - Fix logic error using skb_gso_validate_network_len from Daniel
>
> V6:
> - Took John's advice and dropped BPF_MTU_CHK_RELAX
> - Returned MTU is kept at L3-level (like fib_lookup)
>
> V4: Lot of changes
> - ifindex 0 now use current netdev for MTU lookup
> - rename helper from bpf_mtu_check to bpf_check_mtu
> - fix bug for GSO pkt length (as skb->len is total len)
> - remove __bpf_len_adj_positive, simply allow negative len adj
>
> Signed-off-by: Jesper Dangaard Brouer <bro...@redhat.com>
> ---
> include/uapi/linux/bpf.h | 67 ++++++++++++++++++++++++
> net/core/filter.c | 114
> ++++++++++++++++++++++++++++++++++++++++
> tools/include/uapi/linux/bpf.h | 67 ++++++++++++++++++++++++
> 3 files changed, 248 insertions(+)
>
> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
> index 05bfc8c843dc..f17381a337ec 100644
> --- a/include/uapi/linux/bpf.h
> +++ b/include/uapi/linux/bpf.h
> @@ -3839,6 +3839,61 @@ union bpf_attr {
[...]
> +
> +BPF_CALL_5(bpf_skb_check_mtu, struct sk_buff *, skb,
> + u32, ifindex, u32 *, mtu_len, s32, len_diff, u64, flags)
Maybe worth mentioning in description we expect len_diff < skb->len,
at least I expect that otherwise result may be undefined.
> +{
> + int ret = BPF_MTU_CHK_RET_FRAG_NEEDED;
> + struct net_device *dev = skb->dev;
> + int skb_len, dev_len;
> + int mtu;
Perhaps getting a bit nit-picky here but shouldn't skb_len, dev_len
and mtu all be 'unsigned int'
Then all the types will align. I guess MTUs are small so it
doesn't really matter, but is easier to read IMO.
> +
> + if (unlikely(flags & ~(BPF_MTU_CHK_SEGS)))
> + return -EINVAL;
> +
> + if (unlikely(flags & BPF_MTU_CHK_SEGS && len_diff))
> + return -EINVAL;
> +
> + dev = __dev_via_ifindex(dev, ifindex);
> + if (unlikely(!dev))
> + return -ENODEV;
> +
> + mtu = READ_ONCE(dev->mtu);
> +
> + dev_len = mtu + dev->hard_header_len;
> + skb_len = skb->len + len_diff; /* minus result pass check */
> + if (skb_len <= dev_len) {
If skb_len is unsigned it will be >> dev_len when skb->len < len_diff. I
think its a good idea to throw an error if skb_len calculation goes
negative?
> + ret = BPF_MTU_CHK_RET_SUCCESS;
> + goto out;
> + }
> + /* At this point, skb->len exceed MTU, but as it include length of all
> + * segments, it can still be below MTU. The SKB can possibly get
> + * re-segmented in transmit path (see validate_xmit_skb). Thus, user
> + * must choose if segs are to be MTU checked.
> + */
> + if (skb_is_gso(skb)) {
> + ret = BPF_MTU_CHK_RET_SUCCESS;
> +
> + if (flags & BPF_MTU_CHK_SEGS &&
> + !skb_gso_validate_network_len(skb, mtu))
> + ret = BPF_MTU_CHK_RET_SEGS_TOOBIG;
> + }
> +out:
> + /* BPF verifier guarantees valid pointer */
> + *mtu_len = mtu;
> +
> + return ret;
> +}
> +
> +BPF_CALL_5(bpf_xdp_check_mtu, struct xdp_buff *, xdp,
> + u32, ifindex, u32 *, mtu_len, s32, len_diff, u64, flags)
> +{
> + struct net_device *dev = xdp->rxq->dev;
> + int xdp_len = xdp->data_end - xdp->data;
> + int ret = BPF_MTU_CHK_RET_SUCCESS;
> + int mtu, dev_len;
Same comment about types.
> +
> + /* XDP variant doesn't support multi-buffer segment check (yet) */
> + if (unlikely(flags))
> + return -EINVAL;
> +
> + dev = __dev_via_ifindex(dev, ifindex);
> + if (unlikely(!dev))
> + return -ENODEV;
> +
> + mtu = READ_ONCE(dev->mtu);
> +
> + /* Add L2-header as dev MTU is L3 size */
> + dev_len = mtu + dev->hard_header_len;
> +
> + xdp_len += len_diff; /* minus result pass check */
> + if (xdp_len > dev_len)
> + ret = BPF_MTU_CHK_RET_FRAG_NEEDED;
> +
> + /* BPF verifier guarantees valid pointer */
> + *mtu_len = mtu;
> +
> + return ret;
> +}
Otherwise LGTM.
