diff options
author | Jakub Sitnicki | 2020-04-29 20:11:52 +0200 |
---|---|---|
committer | Daniel Borkmann | 2020-04-29 23:30:59 +0200 |
commit | 64d85290d79c0677edb5a8ee2295b36c022fa5df (patch) | |
tree | b8bc403085811e0d4358be3d4d57ae24a6ff1e2f /kernel/bpf | |
parent | 0b3b9ca3d154486baa08a41cbc62fde67ba8c6c3 (diff) |
bpf: Allow bpf_map_lookup_elem for SOCKMAP and SOCKHASH
White-list map lookup for SOCKMAP/SOCKHASH from BPF. Lookup returns a
pointer to a full socket and acquires a reference if necessary.
To support it we need to extend the verifier to know that:
(1) register storing the lookup result holds a pointer to socket, if
lookup was done on SOCKMAP/SOCKHASH, and that
(2) map lookup on SOCKMAP/SOCKHASH is a reference acquiring operation,
which needs a corresponding reference release with bpf_sk_release.
On sock_map side, lookup handlers exposed via bpf_map_ops now bump
sk_refcnt if socket is reference counted. In turn, bpf_sk_select_reuseport,
the only in-kernel user of SOCKMAP/SOCKHASH ops->map_lookup_elem, was
updated to release the reference.
Sockets fetched from a map can be used in the same way as ones returned by
BPF socket lookup helpers, such as bpf_sk_lookup_tcp. In particular, they
can be used with bpf_sk_assign to direct packets toward a socket on TC
ingress path.
Suggested-by: Lorenz Bauer <lmb@cloudflare.com>
Signed-off-by: Jakub Sitnicki <jakub@cloudflare.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Link: https://lore.kernel.org/bpf/20200429181154.479310-2-jakub@cloudflare.com
Diffstat (limited to 'kernel/bpf')
-rw-r--r-- | kernel/bpf/verifier.c | 45 |
1 files changed, 35 insertions, 10 deletions
diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 2b337e32aa94..70ad009577f8 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -429,11 +429,30 @@ static bool is_release_function(enum bpf_func_id func_id) return func_id == BPF_FUNC_sk_release; } -static bool is_acquire_function(enum bpf_func_id func_id) +static bool may_be_acquire_function(enum bpf_func_id func_id) { return func_id == BPF_FUNC_sk_lookup_tcp || func_id == BPF_FUNC_sk_lookup_udp || - func_id == BPF_FUNC_skc_lookup_tcp; + func_id == BPF_FUNC_skc_lookup_tcp || + func_id == BPF_FUNC_map_lookup_elem; +} + +static bool is_acquire_function(enum bpf_func_id func_id, + const struct bpf_map *map) +{ + enum bpf_map_type map_type = map ? map->map_type : BPF_MAP_TYPE_UNSPEC; + + if (func_id == BPF_FUNC_sk_lookup_tcp || + func_id == BPF_FUNC_sk_lookup_udp || + func_id == BPF_FUNC_skc_lookup_tcp) + return true; + + if (func_id == BPF_FUNC_map_lookup_elem && + (map_type == BPF_MAP_TYPE_SOCKMAP || + map_type == BPF_MAP_TYPE_SOCKHASH)) + return true; + + return false; } static bool is_ptr_cast_function(enum bpf_func_id func_id) @@ -3934,7 +3953,8 @@ static int check_map_func_compatibility(struct bpf_verifier_env *env, func_id != BPF_FUNC_sock_map_update && func_id != BPF_FUNC_map_delete_elem && func_id != BPF_FUNC_msg_redirect_map && - func_id != BPF_FUNC_sk_select_reuseport) + func_id != BPF_FUNC_sk_select_reuseport && + func_id != BPF_FUNC_map_lookup_elem) goto error; break; case BPF_MAP_TYPE_SOCKHASH: @@ -3942,7 +3962,8 @@ static int check_map_func_compatibility(struct bpf_verifier_env *env, func_id != BPF_FUNC_sock_hash_update && func_id != BPF_FUNC_map_delete_elem && func_id != BPF_FUNC_msg_redirect_hash && - func_id != BPF_FUNC_sk_select_reuseport) + func_id != BPF_FUNC_sk_select_reuseport && + func_id != BPF_FUNC_map_lookup_elem) goto error; break; case BPF_MAP_TYPE_REUSEPORT_SOCKARRAY: @@ -4112,7 +4133,7 @@ static bool check_refcount_ok(const struct bpf_func_proto *fn, int func_id) /* A reference acquiring function cannot acquire * another refcounted ptr. */ - if (is_acquire_function(func_id) && count) + if (may_be_acquire_function(func_id) && count) return false; /* We only support one arg being unreferenced at the moment, @@ -4623,7 +4644,7 @@ static int check_helper_call(struct bpf_verifier_env *env, int func_id, int insn if (is_ptr_cast_function(func_id)) { /* For release_reference() */ regs[BPF_REG_0].ref_obj_id = meta.ref_obj_id; - } else if (is_acquire_function(func_id)) { + } else if (is_acquire_function(func_id, meta.map_ptr)) { int id = acquire_reference_state(env, insn_idx); if (id < 0) @@ -6532,12 +6553,16 @@ static void mark_ptr_or_null_reg(struct bpf_func_state *state, if (is_null) { reg->type = SCALAR_VALUE; } else if (reg->type == PTR_TO_MAP_VALUE_OR_NULL) { - if (reg->map_ptr->inner_map_meta) { + const struct bpf_map *map = reg->map_ptr; + + if (map->inner_map_meta) { reg->type = CONST_PTR_TO_MAP; - reg->map_ptr = reg->map_ptr->inner_map_meta; - } else if (reg->map_ptr->map_type == - BPF_MAP_TYPE_XSKMAP) { + reg->map_ptr = map->inner_map_meta; + } else if (map->map_type == BPF_MAP_TYPE_XSKMAP) { reg->type = PTR_TO_XDP_SOCK; + } else if (map->map_type == BPF_MAP_TYPE_SOCKMAP || + map->map_type == BPF_MAP_TYPE_SOCKHASH) { + reg->type = PTR_TO_SOCKET; } else { reg->type = PTR_TO_MAP_VALUE; } |