[PATCH bpf-next V1 0/2] bpf: Updates for BPF-helper bpf_check_mtu

[PATCH bpf-next V1 0/2] bpf: Updates for BPF-helper bpf_check_mtu

[Jesper Dangaard Brouer]

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. The recently
added MTU check helper bpf_check_mtu() should also support this type
of MTU check.

Lets add this feature before merge window, please. This is a followup
to 34b2021cc616 ("bpf: Add BPF-helper for MTU checking").

[1] samples/bpf/xdp_fwd_kern.c
---

Jesper Dangaard Brouer (2):
      bpf: BPF-helper for MTU checking add length input
      selftests/bpf: Tests using bpf_check_mtu BPF-helper input mtu_len param


 include/uapi/linux/bpf.h                           |   17 ++--
 net/core/filter.c                                  |   12 ++-
 tools/testing/selftests/bpf/prog_tests/check_mtu.c |    4 +
 tools/testing/selftests/bpf/progs/test_check_mtu.c |   92 ++++++++++++++++++++
 4 files changed, 117 insertions(+), 8 deletions(-)

--

[PATCH bpf-next V1 1/2] bpf: BPF-helper for MTU checking add length input

[Jesper Dangaard Brouer]

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.

[1] samples/bpf/xdp_fwd_kern.c

Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
---
 include/uapi/linux/bpf.h |   17 +++++++++++------
 net/core/filter.c        |   12 ++++++++++--
 2 files changed, 21 insertions(+), 8 deletions(-)

diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
index 4c24daa43bac..9c8aa50dc8a5 100644
--- a/include/uapi/linux/bpf.h
+++ b/include/uapi/linux/bpf.h
@@ -3850,8 +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.
  *
@@ -3868,6 +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. Remeber 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**\ ()
@@ -3892,11 +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.
diff --git a/net/core/filter.c b/net/core/filter.c
index 7059cf604d94..fcc3bda85960 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -5660,7 +5660,7 @@ 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);
@@ -5670,7 +5670,11 @@ 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;
@@ -5715,6 +5719,10 @@ 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;

[PATCH bpf-next V1 2/2] selftests/bpf: Tests using bpf_check_mtu BPF-helper input mtu_len param

[Jesper Dangaard Brouer]

Add tests that use mtu_len as input parameter in BPF-helper
bpf_check_mtu().

The BPF-helper is avail from both XDP and TC context. Add two tests
per context, one that tests below MTU and one that exceeds the MTU.

Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
---
 tools/testing/selftests/bpf/prog_tests/check_mtu.c |    4 +
 tools/testing/selftests/bpf/progs/test_check_mtu.c |   92 ++++++++++++++++++++
 2 files changed, 96 insertions(+)

diff --git a/tools/testing/selftests/bpf/prog_tests/check_mtu.c b/tools/testing/selftests/bpf/prog_tests/check_mtu.c
index 36af1c138faf..b62a39315336 100644
--- a/tools/testing/selftests/bpf/prog_tests/check_mtu.c
+++ b/tools/testing/selftests/bpf/prog_tests/check_mtu.c
@@ -128,6 +128,8 @@ static void test_check_mtu_xdp(__u32 mtu, __u32 ifindex)
 	test_check_mtu_run_xdp(skel, skel->progs.xdp_use_helper, mtu);
 	test_check_mtu_run_xdp(skel, skel->progs.xdp_exceed_mtu, mtu);
 	test_check_mtu_run_xdp(skel, skel->progs.xdp_minus_delta, mtu);
+	test_check_mtu_run_xdp(skel, skel->progs.xdp_input_len, mtu);
+	test_check_mtu_run_xdp(skel, skel->progs.xdp_input_len_exceed, mtu);
 
 cleanup:
 	test_check_mtu__destroy(skel);
@@ -187,6 +189,8 @@ static void test_check_mtu_tc(__u32 mtu, __u32 ifindex)
 	test_check_mtu_run_tc(skel, skel->progs.tc_exceed_mtu, mtu);
 	test_check_mtu_run_tc(skel, skel->progs.tc_exceed_mtu_da, mtu);
 	test_check_mtu_run_tc(skel, skel->progs.tc_minus_delta, mtu);
+	test_check_mtu_run_tc(skel, skel->progs.tc_input_len, mtu);
+	test_check_mtu_run_tc(skel, skel->progs.tc_input_len_exceed, mtu);
 cleanup:
 	test_check_mtu__destroy(skel);
 }
diff --git a/tools/testing/selftests/bpf/progs/test_check_mtu.c b/tools/testing/selftests/bpf/progs/test_check_mtu.c
index b7787b43f9db..c4a9bae96e75 100644
--- a/tools/testing/selftests/bpf/progs/test_check_mtu.c
+++ b/tools/testing/selftests/bpf/progs/test_check_mtu.c
@@ -105,6 +105,54 @@ int xdp_minus_delta(struct xdp_md *ctx)
 	return retval;
 }
 
+SEC("xdp")
+int xdp_input_len(struct xdp_md *ctx)
+{
+	int retval = XDP_PASS; /* Expected retval on successful test */
+	void *data_end = (void *)(long)ctx->data_end;
+	void *data = (void *)(long)ctx->data;
+	__u32 ifindex = GLOBAL_USER_IFINDEX;
+	__u32 data_len = data_end - data;
+
+	/* API allow user give length to check as input via mtu_len param,
+	 * resulting MTU value is still output in mtu_len param after call.
+	 *
+	 * Input len is L3, like MTU and iph->tot_len.
+	 * Remember XDP data_len is L2.
+	 */
+	__u32 mtu_len = data_len - ETH_HLEN;
+
+	if (bpf_check_mtu(ctx, ifindex, &mtu_len, 0, 0))
+		retval = XDP_ABORTED;
+
+	global_bpf_mtu_xdp = mtu_len;
+	return retval;
+}
+
+SEC("xdp")
+int xdp_input_len_exceed(struct xdp_md *ctx)
+{
+	int retval = XDP_ABORTED; /* Fail */
+	__u32 ifindex = GLOBAL_USER_IFINDEX;
+	int err;
+
+	/* API allow user give length to check as input via mtu_len param,
+	 * resulting MTU value is still output in mtu_len param after call.
+	 *
+	 * Input length value is L3 size like MTU.
+	 */
+	__u32 mtu_len = GLOBAL_USER_MTU;
+
+	mtu_len += 1; /* Exceed with 1 */
+
+	err = bpf_check_mtu(ctx, ifindex, &mtu_len, 0, 0);
+	if (err == BPF_MTU_CHK_RET_FRAG_NEEDED)
+		retval = XDP_PASS ; /* Success in exceeding MTU check */
+
+	global_bpf_mtu_xdp = mtu_len;
+	return retval;
+}
+
 SEC("classifier")
 int tc_use_helper(struct __sk_buff *ctx)
 {
@@ -196,3 +244,47 @@ int tc_minus_delta(struct __sk_buff *ctx)
 	global_bpf_mtu_xdp = mtu_len;
 	return retval;
 }
+
+SEC("classifier")
+int tc_input_len(struct __sk_buff *ctx)
+{
+	int retval = BPF_OK; /* Expected retval on successful test */
+	__u32 ifindex = GLOBAL_USER_IFINDEX;
+
+	/* API allow user give length to check as input via mtu_len param,
+	 * resulting MTU value is still output in mtu_len param after call.
+	 *
+	 * Input length value is L3 size.
+	 */
+	__u32 mtu_len = GLOBAL_USER_MTU;
+
+	if (bpf_check_mtu(ctx, ifindex, &mtu_len, 0, 0))
+		retval = BPF_DROP;
+
+	global_bpf_mtu_xdp = mtu_len;
+	return retval;
+}
+
+SEC("classifier")
+int tc_input_len_exceed(struct __sk_buff *ctx)
+{
+	int retval = BPF_DROP; /* Fail */
+	__u32 ifindex = GLOBAL_USER_IFINDEX;
+	int err;
+
+	/* API allow user give length to check as input via mtu_len param,
+	 * resulting MTU value is still output in mtu_len param after call.
+	 *
+	 * Input length value is L3 size like MTU.
+	 */
+	__u32 mtu_len = GLOBAL_USER_MTU;
+
+	mtu_len += 1; /* Exceed with 1 */
+
+	err = bpf_check_mtu(ctx, ifindex, &mtu_len, 0, 0);
+	if (err == BPF_MTU_CHK_RET_FRAG_NEEDED)
+		retval = BPF_OK; /* Success in exceeding MTU check */
+
+	global_bpf_mtu_xdp = mtu_len;
+	return retval;
+}

RE: [PATCH bpf-next V1 1/2] bpf: BPF-helper for MTU checking add length input

[John Fastabend]

Jesper Dangaard Brouer wrote:
> 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.
> 
> [1] samples/bpf/xdp_fwd_kern.c
> 
> Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
> ---
>  include/uapi/linux/bpf.h |   17 +++++++++++------
>  net/core/filter.c        |   12 ++++++++++--
>  2 files changed, 21 insertions(+), 8 deletions(-)
> 
> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
> index 4c24daa43bac..9c8aa50dc8a5 100644
> --- a/include/uapi/linux/bpf.h
> +++ b/include/uapi/linux/bpf.h
> @@ -3850,8 +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.
>   *
> @@ -3868,6 +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. Remeber XDP and TC length operate at L2, but
                                 ^^^^^^^
typo, Remember

Acked-by: John Fastabend <john.fastabend@gmail.com>

RE: [PATCH bpf-next V1 2/2] selftests/bpf: Tests using bpf_check_mtu BPF-helper input mtu_len param

[John Fastabend]

Jesper Dangaard Brouer wrote:
> Add tests that use mtu_len as input parameter in BPF-helper
> bpf_check_mtu().
> 
> The BPF-helper is avail from both XDP and TC context. Add two tests
> per context, one that tests below MTU and one that exceeds the MTU.
> 
> Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
> ---

LGTM.

Acked-by: John Fastabend <john.fastabend@gmail.com>