RE: [Patch bpf-next v1 1/4] tcp: introduce tcp_read_skb()

[John Fastabend <john.fastabend@gmail.com>]

Cong Wang wrote:
> From: Cong Wang <cong.wang@bytedance.com>
> 
> This patch inroduces tcp_read_skb() based on tcp_read_sock(),
> a preparation for the next patch which actually introduces
> a new sock ops.
> 
> TCP is special here, because it has tcp_read_sock() which is
> mainly used by splice(). tcp_read_sock() supports partial read
> and arbitrary offset, neither of them is needed for sockmap.
> 
> Cc: Eric Dumazet <edumazet@google.com>
> Cc: John Fastabend <john.fastabend@gmail.com>
> Cc: Daniel Borkmann <daniel@iogearbox.net>
> Cc: Jakub Sitnicki <jakub@cloudflare.com>
> Signed-off-by: Cong Wang <cong.wang@bytedance.com>
> ---

Thanks for doing this Cong comment/question inline.

[...]

> +int tcp_read_skb(struct sock *sk, read_descriptor_t *desc,
> +		 sk_read_actor_t recv_actor)
> +{
> +	struct sk_buff *skb;
> +	struct tcp_sock *tp = tcp_sk(sk);
> +	u32 seq = tp->copied_seq;
> +	u32 offset;
> +	int copied = 0;
> +
> +	if (sk->sk_state == TCP_LISTEN)
> +		return -ENOTCONN;
> +	while ((skb = tcp_recv_skb(sk, seq, &offset, true)) != NULL) {

I'm trying to see why we might have an offset here if we always
consume the entire skb. There is a comment in tcp_recv_skb around
GRO packets, but not clear how this applies here if it does at all
to me yet. Will read a bit more I guess.

If the offset can be >0 than we also need to fix the recv_actor to
account for the extra offset in the skb. As is the bpf prog might
see duplicate data. This is a problem on the stream parser now.

Then another fallout is if offset is zero than we could just do
a skb_dequeue here and skip the tcp_recv_skb bool flag addition
and upate.

I'll continue reading after a few other things I need to get
sorted this afternoon, but maybe you have the answer on hand.

> +		if (offset < skb->len) {
> +			int used;
> +			size_t len;
> +
> +			len = skb->len - offset;
> +			used = recv_actor(desc, skb, offset, len);
> +			if (used <= 0) {
> +				if (!copied)
> +					copied = used;
> +				break;
> +			}
> +			if (WARN_ON_ONCE(used > len))
> +				used = len;
> +			seq += used;
> +			copied += used;
> +			offset += used;
> +
> +			if (offset != skb->len)
> +				continue;
> +		}
> +		if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN) {
> +			kfree_skb(skb);
> +			++seq;
> +			break;
> +		}
> +		kfree_skb(skb);
> +		if (!desc->count)
> +			break;
> +		WRITE_ONCE(tp->copied_seq, seq);
> +	}
> +	WRITE_ONCE(tp->copied_seq, seq);
> +
> +	tcp_rcv_space_adjust(sk);
> +
> +	/* Clean up data we have read: This will do ACK frames. */
> +	if (copied > 0)
> +		tcp_cleanup_rbuf(sk, copied);
> +
> +	return copied;
> +}
> +EXPORT_SYMBOL(tcp_read_skb);
> +