bpf: BPF-helper for MTU checking add length input

The FIB lookup example[1] show how the IP-header field tot_len
(iph->tot_len) is used as input to perform the MTU check.

This patch extend the BPF-helper bpf_check_mtu() with the same ability
to provide the length as user parameter input, via mtu_len parameter.

This still needs to be done before the bpf_check_mtu() helper API
becomes frozen.

  [1] samples/bpf/xdp_fwd_kern.c

Fixes: 34b2021c ("bpf: Add BPF-helper for MTU checking")
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Link: https://lore.kernel.org/bpf/161521555850.3515614.6533850861569774444.stgit@firesoul

include/uapi/linux/bpf.h

@@ -3850,7 +3850,7 @@ union bpf_attr {
  *
  * long bpf_check_mtu(void *ctx, u32 ifindex, u32 *mtu_len, s32 len_diff, u64 flags)
  *	Description
- *		Check ctx packet size against exceeding MTU of net device (based
+ *		Check packet size against exceeding MTU of net device (based
  *		on *ifindex*).  This helper will likely be used in combination
  *		with helpers that adjust/change the packet size.
  *
@@ -3867,6 +3867,14 @@ union bpf_attr {
  *		against the current net device.  This is practical if this isn't
  *		used prior to redirect.
  *
+ *		On input *mtu_len* must be a valid pointer, else verifier will
+ *		reject BPF program.  If the value *mtu_len* is initialized to
+ *		zero then the ctx packet size is use.  When value *mtu_len* is
+ *		provided as input this specify the L3 length that the MTU check
+ *		is done against. Remember XDP and TC length operate at L2, but
+ *		this value is L3 as this correlate to MTU and IP-header tot_len
+ *		values which are L3 (similar behavior as bpf_fib_lookup).
+ *
  *		The Linux kernel route table can configure MTUs on a more
  *		specific per route level, which is not provided by this helper.
  *		For route level MTU checks use the **bpf_fib_lookup**\ ()
@@ -3891,11 +3899,9 @@ union bpf_attr {
  *
  *		On return *mtu_len* pointer contains the MTU value of the net
  *		device.  Remember the net device configured MTU is the L3 size,
- *		which is returned here and XDP and TX length operate at L2.
+ *		which is returned here and XDP and TC length operate at L2.
  *		Helper take this into account for you, but remember when using
- *		MTU value in your BPF-code.  On input *mtu_len* must be a valid
- *		pointer and be initialized (to zero), else verifier will reject
- *		BPF program.
+ *		MTU value in your BPF-code.
  *
  *	Return
  *		* 0 on success, and populate MTU value in *mtu_len* pointer.

net/core/filter.c

@@ -5658,7 +5658,7 @@ BPF_CALL_5(bpf_skb_check_mtu, struct sk_buff *, skb,
 	if (unlikely(flags & ~(BPF_MTU_CHK_SEGS)))
 		return -EINVAL;
 
-	if (unlikely(flags & BPF_MTU_CHK_SEGS && len_diff))
+	if (unlikely(flags & BPF_MTU_CHK_SEGS && (len_diff || *mtu_len)))
 		return -EINVAL;
 
 	dev = __dev_via_ifindex(dev, ifindex);
@@ -5668,7 +5668,11 @@ BPF_CALL_5(bpf_skb_check_mtu, struct sk_buff *, skb,
 	mtu = READ_ONCE(dev->mtu);
 
 	dev_len = mtu + dev->hard_header_len;
-	skb_len = skb->len + len_diff; /* minus result pass check */
+
+	/* If set use *mtu_len as input, L3 as iph->tot_len (like fib_lookup) */
+	skb_len = *mtu_len ? *mtu_len + dev->hard_header_len : skb->len;
+
+	skb_len += len_diff; /* minus result pass check */
 	if (skb_len <= dev_len) {
 		ret = BPF_MTU_CHK_RET_SUCCESS;
 		goto out;
@@ -5713,6 +5717,10 @@ BPF_CALL_5(bpf_xdp_check_mtu, struct xdp_buff *, xdp,
 	/* Add L2-header as dev MTU is L3 size */
 	dev_len = mtu + dev->hard_header_len;
 
+	/* Use *mtu_len as input, L3 as iph->tot_len (like fib_lookup) */
+	if (*mtu_len)
+		xdp_len = *mtu_len + dev->hard_header_len;
+
 	xdp_len += len_diff; /* minus result pass check */
 	if (xdp_len > dev_len)
 		ret = BPF_MTU_CHK_RET_FRAG_NEEDED;