On Tue, Sep 1, 2020 at 3:33 AM Lorenz Bauer <lmb@xxxxxxxxxxxxxx> wrote:
>
> Add bpf_iter support for sockmap / sockhash, based on the bpf_sk_storage and
> hashtable implementation. sockmap and sockhash share the same iteration
> context: a pointer to an arbitrary key and a pointer to a socket. Both
> pointers may be NULL, and so BPF has to perform a NULL check before accessing
> them. Technically it's not possible for sockhash iteration to yield a NULL
> socket, but we ignore this to be able to use a single iteration point.
>
> Iteration will visit all keys that remain unmodified during the lifetime of
> the iterator. It may or may not visit newly added ones.
>
> Signed-off-by: Lorenz Bauer <lmb@xxxxxxxxxxxxxx>
> ---
> net/core/sock_map.c | 283 ++++++++++++++++++++++++++++++++++++++++++++
> 1 file changed, 283 insertions(+)
>
> diff --git a/net/core/sock_map.c b/net/core/sock_map.c
> index ffdf94a30c87..4767f9df2b8b 100644
> --- a/net/core/sock_map.c
> +++ b/net/core/sock_map.c
> @@ -703,6 +703,114 @@ const struct bpf_func_proto bpf_msg_redirect_map_proto = {
> .arg4_type = ARG_ANYTHING,
> };
>
> +struct sock_map_seq_info {
> + struct bpf_map *map;
> + struct sock *sk;
> + u32 index;
> +};
> +
> +struct bpf_iter__sockmap {
> + __bpf_md_ptr(struct bpf_iter_meta *, meta);
> + __bpf_md_ptr(struct bpf_map *, map);
> + __bpf_md_ptr(void *, key);
For sockhash, the key can be of an arbitrary size, right? Should the
key_size be part of bpf_iter__sockmap then?
> + __bpf_md_ptr(struct bpf_sock *, sk);
> +};
> +
[...]
