On 03/12/2018 08:23 PM, John Fastabend wrote:
> A single sendmsg or sendfile system call can contain multiple logical
> messages that a BPF program may want to read and apply a verdict. But,
> without an apply_bytes helper any verdict on the data applies to all
> bytes in the sendmsg/sendfile. Alternatively, a BPF program may only
> care to read the first N bytes of a msg. If the payload is large say
> MB or even GB setting up and calling the BPF program repeatedly for
> all bytes, even though the verdict is already known, creates
> unnecessary overhead.
>
> To allow BPF programs to control how many bytes a given verdict
> applies to we implement a bpf_msg_apply_bytes() helper. When called
> from within a BPF program this sets a counter, internal to the
> BPF infrastructure, that applies the last verdict to the next N
> bytes. If the N is smaller than the current data being processed
> from a sendmsg/sendfile call, the first N bytes will be sent and
> the BPF program will be re-run with start_data pointing to the N+1
> byte. If N is larger than the current data being processed the
> BPF verdict will be applied to multiple sendmsg/sendfile calls
> until N bytes are consumed.
>
> Note1 if a socket closes with apply_bytes counter non-zero this
> is not a problem because data is not being buffered for N bytes
> and is sent as its received.
>
> Note2 if this is operating in the sendpage context the data
> pointers may be zeroed after this call if the apply walks beyond
> a msg_pull_data() call specified data range. (helper implemented
> shortly in this series).
>
> Signed-off-by: John Fastabend <john.fastab...@gmail.com>
> ---
> include/uapi/linux/bpf.h | 3 ++-
> net/core/filter.c | 16 ++++++++++++++++
> 2 files changed, 18 insertions(+), 1 deletion(-)
>
> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
> index b8275f0..e50c61f 100644
> --- a/include/uapi/linux/bpf.h
> +++ b/include/uapi/linux/bpf.h
> @@ -769,7 +769,8 @@ enum bpf_attach_type {
> FN(getsockopt), \
> FN(override_return), \
> FN(sock_ops_cb_flags_set), \
> - FN(msg_redirect_map),
> + FN(msg_redirect_map), \
> + FN(msg_apply_bytes),
>
> /* integer value in 'imm' field of BPF_CALL instruction selects which helper
> * function eBPF program intends to call
> diff --git a/net/core/filter.c b/net/core/filter.c
> index 314c311..df2a8f4 100644
> --- a/net/core/filter.c
> +++ b/net/core/filter.c
> @@ -1928,6 +1928,20 @@ struct sock *do_msg_redirect_map(struct sk_msg_buff
> *msg)
> .arg4_type = ARG_ANYTHING,
> };
>
> +BPF_CALL_2(bpf_msg_apply_bytes, struct sk_msg_buff *, msg, u64, bytes)
> +{
> + msg->apply_bytes = bytes;
Here in bpf_msg_apply_bytes() but also in bpf_msg_cork_bytes() the signature
is u64, but in struct sk_msg_buff and struct smap_psock it's type int, so
user provided u64 will make these negative. Is there a reason to have this
allow a negative value and not being of type u32 everywhere?
> + return 0;
> +}
> +
> +static const struct bpf_func_proto bpf_msg_apply_bytes_proto = {
> + .func = bpf_msg_apply_bytes,
> + .gpl_only = false,
> + .ret_type = RET_INTEGER,
> + .arg1_type = ARG_PTR_TO_CTX,
> + .arg2_type = ARG_ANYTHING,
> +};
> +
> BPF_CALL_1(bpf_get_cgroup_classid, const struct sk_buff *, skb)
> {
> return task_get_classid(skb);
> @@ -3634,6 +3648,8 @@ static const struct bpf_func_proto
> *sk_msg_func_proto(enum bpf_func_id func_id)
> switch (func_id) {
> case BPF_FUNC_msg_redirect_map:
> return &bpf_msg_redirect_map_proto;
> + case BPF_FUNC_msg_apply_bytes:
> + return &bpf_msg_apply_bytes_proto;
> default:
> return bpf_base_func_proto(func_id);
> }
>
