From 18ebe16d10496db795a9930c320c45653207a548 Mon Sep 17 00:00:00 2001 From: John Fastabend Date: Thu, 1 Oct 2020 18:09:52 -0700 Subject: bpf, sockmap: Add skb_adjust_room to pop bytes off ingress payload This implements a new helper skb_adjust_room() so users can push/pop extra bytes from a BPF_SK_SKB_STREAM_VERDICT program. Some protocols may include headers and other information that we may not want to include when doing a redirect from a BPF_SK_SKB_STREAM_VERDICT program. One use case is to redirect TLS packets into a receive socket that doesn't expect TLS data. In TLS case the first 13B or so contain the protocol header. With KTLS the payload is decrypted so we should be able to redirect this to a receiving socket, but the receiving socket may not be expecting to receive a TLS header and discard the data. Using the above helper we can pop the header off and put an appropriate header on the payload. This allows for creating a proxy between protocols without extra hops through the stack or userspace. So in order to fix this case add skb_adjust_room() so users can strip the header. After this the user can strip the header and an unmodified receiver thread will work correctly when data is redirected into the ingress path of a sock. Signed-off-by: John Fastabend Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/160160099197.7052.8443193973242831692.stgit@john-Precision-5820-Tower --- net/core/filter.c | 46 ++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 46 insertions(+) diff --git a/net/core/filter.c b/net/core/filter.c index 3fb6adad1957..05df73780dd3 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -76,6 +76,7 @@ #include #include #include +#include static const struct bpf_func_proto * bpf_sk_base_func_proto(enum bpf_func_id func_id); @@ -3479,6 +3480,48 @@ static u32 __bpf_skb_max_len(const struct sk_buff *skb) SKB_MAX_ALLOC; } +BPF_CALL_4(sk_skb_adjust_room, struct sk_buff *, skb, s32, len_diff, + u32, mode, u64, flags) +{ + u32 len_diff_abs = abs(len_diff); + bool shrink = len_diff < 0; + int ret = 0; + + if (unlikely(flags || mode)) + return -EINVAL; + if (unlikely(len_diff_abs > 0xfffU)) + return -EFAULT; + + if (!shrink) { + ret = skb_cow(skb, len_diff); + if (unlikely(ret < 0)) + return ret; + __skb_push(skb, len_diff_abs); + memset(skb->data, 0, len_diff_abs); + } else { + if (unlikely(!pskb_may_pull(skb, len_diff_abs))) + return -ENOMEM; + __skb_pull(skb, len_diff_abs); + } + bpf_compute_data_end_sk_skb(skb); + if (tls_sw_has_ctx_rx(skb->sk)) { + struct strp_msg *rxm = strp_msg(skb); + + rxm->full_len += len_diff; + } + return ret; +} + +static const struct bpf_func_proto sk_skb_adjust_room_proto = { + .func = sk_skb_adjust_room, + .gpl_only = false, + .ret_type = RET_INTEGER, + .arg1_type = ARG_PTR_TO_CTX, + .arg2_type = ARG_ANYTHING, + .arg3_type = ARG_ANYTHING, + .arg4_type = ARG_ANYTHING, +}; + BPF_CALL_4(bpf_skb_adjust_room, struct sk_buff *, skb, s32, len_diff, u32, mode, u64, flags) { @@ -6745,6 +6788,7 @@ bool bpf_helper_changes_pkt_data(void *func) func == bpf_skb_change_tail || func == sk_skb_change_tail || func == bpf_skb_adjust_room || + func == sk_skb_adjust_room || func == bpf_skb_pull_data || func == sk_skb_pull_data || func == bpf_clone_redirect || @@ -7218,6 +7262,8 @@ sk_skb_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) return &sk_skb_change_tail_proto; case BPF_FUNC_skb_change_head: return &sk_skb_change_head_proto; + case BPF_FUNC_skb_adjust_room: + return &sk_skb_adjust_room_proto; case BPF_FUNC_get_socket_cookie: return &bpf_get_socket_cookie_proto; case BPF_FUNC_get_socket_uid: -- cgit From 91274ca535185813cd5eebfe5a4d3344750c74e3 Mon Sep 17 00:00:00 2001 From: John Fastabend Date: Thu, 1 Oct 2020 18:10:09 -0700 Subject: bpf, sockmap: Update selftests to use skb_adjust_room Instead of working around TLS headers in sockmap selftests use the new skb_adjust_room helper. This allows us to avoid special casing the receive side to skip headers. Signed-off-by: John Fastabend Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/160160100932.7052.3646935243867660528.stgit@john-Precision-5820-Tower --- .../selftests/bpf/progs/test_sockmap_kern.h | 34 ++++++++++++++++------ tools/testing/selftests/bpf/test_sockmap.c | 27 ++++------------- 2 files changed, 31 insertions(+), 30 deletions(-) diff --git a/tools/testing/selftests/bpf/progs/test_sockmap_kern.h b/tools/testing/selftests/bpf/progs/test_sockmap_kern.h index 3dca4c2e2418..1858435de7aa 100644 --- a/tools/testing/selftests/bpf/progs/test_sockmap_kern.h +++ b/tools/testing/selftests/bpf/progs/test_sockmap_kern.h @@ -131,39 +131,55 @@ int bpf_prog2(struct __sk_buff *skb) } -SEC("sk_skb3") -int bpf_prog3(struct __sk_buff *skb) +static inline void bpf_write_pass(struct __sk_buff *skb, int offset) { - const int one = 1; - int err, *f, ret = SK_PASS; + int err = bpf_skb_pull_data(skb, 6 + offset); void *data_end; char *c; - err = bpf_skb_pull_data(skb, 19); if (err) - goto tls_out; + return; c = (char *)(long)skb->data; data_end = (void *)(long)skb->data_end; - if (c + 18 < data_end) - memcpy(&c[13], "PASS", 4); + if (c + 5 + offset < data_end) + memcpy(c + offset, "PASS", 4); +} + +SEC("sk_skb3") +int bpf_prog3(struct __sk_buff *skb) +{ + int err, *f, ret = SK_PASS; + const int one = 1; + f = bpf_map_lookup_elem(&sock_skb_opts, &one); if (f && *f) { __u64 flags = 0; ret = 0; flags = *f; + + err = bpf_skb_adjust_room(skb, -13, 0, 0); + if (err) + return SK_DROP; + err = bpf_skb_adjust_room(skb, 4, 0, 0); + if (err) + return SK_DROP; + bpf_write_pass(skb, 0); #ifdef SOCKMAP return bpf_sk_redirect_map(skb, &tls_sock_map, ret, flags); #else return bpf_sk_redirect_hash(skb, &tls_sock_map, &ret, flags); #endif } - f = bpf_map_lookup_elem(&sock_skb_opts, &one); if (f && *f) ret = SK_DROP; + err = bpf_skb_adjust_room(skb, 4, 0, 0); + if (err) + return SK_DROP; + bpf_write_pass(skb, 13); tls_out: return ret; } diff --git a/tools/testing/selftests/bpf/test_sockmap.c b/tools/testing/selftests/bpf/test_sockmap.c index 9b6fb00dc7a0..5cf45455de42 100644 --- a/tools/testing/selftests/bpf/test_sockmap.c +++ b/tools/testing/selftests/bpf/test_sockmap.c @@ -518,28 +518,13 @@ static int msg_verify_data(struct msghdr *msg, int size, int chunk_sz) if (i == 0 && txmsg_ktls_skb) { if (msg->msg_iov[i].iov_len < 4) return -EIO; - if (txmsg_ktls_skb_redir) { - if (memcmp(&d[13], "PASS", 4) != 0) { - fprintf(stderr, - "detected redirect ktls_skb data error with skb ingress update @iov[%i]:%i \"%02x %02x %02x %02x\" != \"PASS\"\n", i, 0, d[13], d[14], d[15], d[16]); - return -EIO; - } - d[13] = 0; - d[14] = 1; - d[15] = 2; - d[16] = 3; - j = 13; - } else if (txmsg_ktls_skb) { - if (memcmp(d, "PASS", 4) != 0) { - fprintf(stderr, - "detected ktls_skb data error with skb ingress update @iov[%i]:%i \"%02x %02x %02x %02x\" != \"PASS\"\n", i, 0, d[0], d[1], d[2], d[3]); - return -EIO; - } - d[0] = 0; - d[1] = 1; - d[2] = 2; - d[3] = 3; + if (memcmp(d, "PASS", 4) != 0) { + fprintf(stderr, + "detected skb data error with skb ingress update @iov[%i]:%i \"%02x %02x %02x %02x\" != \"PASS\"\n", + i, 0, d[0], d[1], d[2], d[3]); + return -EIO; } + j = 4; /* advance index past PASS header */ } for (; j < msg->msg_iov[i].iov_len && size; j++) { -- cgit