On Fri, Nov 13, 2020 at 12:27 AM CET, John Fastabend wrote:
> If the skb_verdict_prog redirects an skb knowingly to itself, fix your
> BPF program this is not optimal and an abuse of the API please use
> SK_PASS. That said there may be cases, such as socket load balancing,
> where picking the socket is hashed based or otherwise picks the same
> socket it was received on in some rare cases. If this happens we don't
> want to confuse userspace giving them an EAGAIN error if we can avoid
> it.
>
> To avoid double accounting in these cases. At the moment even if the
> skb has already been charged against the sockets rcvbuf and forward
> alloc we check it again and do set_owner_r() causing it to be orphaned
> and recharged. For one this is useless work, but more importantly we
> can have a case where the skb could be put on the ingress queue, but
> because we are under memory pressure we return EAGAIN. The trouble
> here is the skb has already been accounted for so any rcvbuf checks
> include the memory associated with the packet already. This rolls
> up and can result in unecessary EAGAIN errors in userspace read()
> calls.
>
> Fix by doing an unlikely check and skipping checks if skb->sk == sk.
>
> Fixes: 51199405f9672 ("bpf: skb_verdict, support SK_PASS on RX BPF path")
> Signed-off-by: John Fastabend <john.fastabend@xxxxxxxxx>
> ---
> net/core/skmsg.c | 17 +++++++++++------
> 1 file changed, 11 insertions(+), 6 deletions(-)
>
> diff --git a/net/core/skmsg.c b/net/core/skmsg.c
> index 9aed5a2c7c5b..f747ee341fe8 100644
> --- a/net/core/skmsg.c
> +++ b/net/core/skmsg.c
> @@ -404,11 +404,13 @@ static struct sk_msg *sk_psock_create_ingress_msg(struct sock *sk,
> {
> struct sk_msg *msg;
>
> - if (atomic_read(&sk->sk_rmem_alloc) > sk->sk_rcvbuf)
> - return NULL;
> + if (likely(skb->sk != sk)) {
> + if (atomic_read(&sk->sk_rmem_alloc) > sk->sk_rcvbuf)
> + return NULL;
>
> - if (!sk_rmem_schedule(sk, skb, skb->truesize))
> - return NULL;
> + if (!sk_rmem_schedule(sk, skb, skb->truesize))
> + return NULL;
> + }
>
> msg = kzalloc(sizeof(*msg), __GFP_NOWARN | GFP_ATOMIC);
> if (unlikely(!msg))
> @@ -455,9 +457,12 @@ static int sk_psock_skb_ingress(struct sk_psock *psock, struct sk_buff *skb)
> * the BPF program was run initiating the redirect to the socket
> * we will eventually receive this data on. The data will be released
> * from skb_consume found in __tcp_bpf_recvmsg() after its been copied
> - * into user buffers.
> + * into user buffers. If we are receiving on the same sock skb->sk is
> + * already assigned, skip memory accounting and owner transition seeing
> + * it already set correctly.
> */
> - skb_set_owner_r(skb, sk);
> + if (likely(skb->sk != sk))
> + skb_set_owner_r(skb, sk);
> return sk_psock_skb_ingress_enqueue(skb, psock, sk, msg);
> }
>
I think all the added checks boil down to having:
struct sock *sk = psock->sk;
if (unlikely(skb->sk == sk))
return sk_psock_skb_ingress_self(psock, skb);
... on entry to sk_psock_skb_ingress().
