Re: [PATCH net v3] tcp: check skb is non-NULL in tcp_rto_delta_us()

[Neal Cardwell]

On Thu, Sep 19, 2024 at 5:05 AM Paolo Abeni <pabeni@xxxxxxxxxx> wrote:
>
> On 9/10/24 21:08, Josh Hunt wrote:
> > diff --git a/include/net/tcp.h b/include/net/tcp.h
> > index 2aac11e7e1cc..196c148fce8a 100644
> > --- a/include/net/tcp.h
> > +++ b/include/net/tcp.h
> > @@ -2434,9 +2434,26 @@ static inline s64 tcp_rto_delta_us(const struct sock *sk)
> >   {
> >       const struct sk_buff *skb = tcp_rtx_queue_head(sk);
> >       u32 rto = inet_csk(sk)->icsk_rto;
> > -     u64 rto_time_stamp_us = tcp_skb_timestamp_us(skb) + jiffies_to_usecs(rto);
> >
> > -     return rto_time_stamp_us - tcp_sk(sk)->tcp_mstamp;
> > +     if (likely(skb)) {
> > +             u64 rto_time_stamp_us = tcp_skb_timestamp_us(skb) + jiffies_to_usecs(rto);
> > +
> > +             return rto_time_stamp_us - tcp_sk(sk)->tcp_mstamp;
> > +     } else {
> > +             WARN_ONCE(1,
> > +                     "rtx queue emtpy: "
> > +                     "out:%u sacked:%u lost:%u retrans:%u "
> > +                     "tlp_high_seq:%u sk_state:%u ca_state:%u "
> > +                     "advmss:%u mss_cache:%u pmtu:%u\n",
> > +                     tcp_sk(sk)->packets_out, tcp_sk(sk)->sacked_out,
> > +                     tcp_sk(sk)->lost_out, tcp_sk(sk)->retrans_out,
> > +                     tcp_sk(sk)->tlp_high_seq, sk->sk_state,
> > +                     inet_csk(sk)->icsk_ca_state,
> > +                     tcp_sk(sk)->advmss, tcp_sk(sk)->mss_cache,
> > +                     inet_csk(sk)->icsk_pmtu_cookie);
>
> As the underlying issue here share the same root cause as the one
> covered by the WARN_ONCE() in tcp_send_loss_probe(), I'm wondering if it
> would make sense do move the info dumping in a common helper, so that we
> get the verbose warning on either cases.

That's a good idea. It would be nice to move the info dumping into a
common helper and use it from both tcp_rto_delta_us() and
tcp_send_loss_probe(), if Josh is open to that.

Thanks,
neal