diff options
| author |   Jesper Dangaard Brouer <brouer@redhat.com> | 2021-02-09 14:38:34 +0100 |
|---|---|---|
| committer |   Daniel Borkmann <daniel@iogearbox.net> | 2021-02-13 01:15:28 +0100 |
| commit | 6b8838be7e21edb620099e01eb040c21c5190494 (patch) | |
| tree | f81c32dc4ec665afc308c15a97371a99a7fd3ef4 | |
| parent | bpf: Drop MTU check when doing TC-BPF redirect to ingress (diff) | |
| download | wireguard-linux-6b8838be7e21edb620099e01eb040c21c5190494.tar.xz wireguard-linux-6b8838be7e21edb620099e01eb040c21c5190494.zip | |
selftests/bpf: Use bpf_check_mtu in selftest test_cls_redirect
This demonstrate how bpf_check_mtu() helper can easily be used together
with bpf_skb_adjust_room() helper, prior to doing size adjustment, as
delta argument is already setup.
Hint: This specific test can be selected like this:
./test_progs -t cls_redirect
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/161287791481.790810.4444271170546646080.stgit@firesoul
| -rw-r--r-- | tools/testing/selftests/bpf/progs/test_cls_redirect.c | 7 |
1 files changed, 7 insertions, 0 deletions
diff --git a/tools/testing/selftests/bpf/progs/test_cls_redirect.c b/tools/testing/selftests/bpf/progs/test_cls_redirect.c index c9f8464996ea..3c1e042962e6 100644 --- a/tools/testing/selftests/bpf/progs/test_cls_redirect.c +++ b/tools/testing/selftests/bpf/progs/test_cls_redirect.c @@ -70,6 +70,7 @@ typedef struct { uint64_t errors_total_encap_adjust_failed; uint64_t errors_total_encap_buffer_too_small; uint64_t errors_total_redirect_loop; + uint64_t errors_total_encap_mtu_violate; } metrics_t; typedef enum { @@ -407,6 +408,7 @@ static INLINING ret_t forward_with_gre(struct __sk_buff *skb, encap_headers_t *e payload_off - sizeof(struct ethhdr) - sizeof(struct iphdr); int32_t delta = sizeof(struct gre_base_hdr) - encap_overhead; uint16_t proto = ETH_P_IP; + uint32_t mtu_len = 0; /* Loop protection: the inner packet's TTL is decremented as a safeguard * against any forwarding loop. As the only interesting field is the TTL @@ -479,6 +481,11 @@ static INLINING ret_t forward_with_gre(struct __sk_buff *skb, encap_headers_t *e } } + if (bpf_check_mtu(skb, skb->ifindex, &mtu_len, delta, 0)) { + metrics->errors_total_encap_mtu_violate++; + return TC_ACT_SHOT; + } + if (bpf_skb_adjust_room(skb, delta, BPF_ADJ_ROOM_NET, BPF_F_ADJ_ROOM_FIXED_GSO | BPF_F_ADJ_ROOM_NO_CSUM_RESET) || |