netdev - Re: [PATCH net 1/1] tcp: check skb is non-NULL in tcp_rto_delta

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <CANn89i+jT0E_N4k=ciw7XvJXLH15rA=8qexRi=7D3YPo5=ZoqA@mail.gmail.com>
Date: Fri, 23 Aug 2024 08:42:01 +0200
From: Eric Dumazet <edumazet@...gle.com>
To: Josh Hunt <johunt@...mai.com>, Neal Cardwell <ncardwell@...gle.com>
Cc: davem@...emloft.net, kuba@...nel.org, pabeni@...hat.com, 
	netdev@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH net 1/1] tcp: check skb is non-NULL in tcp_rto_delta_us()

On Fri, Aug 23, 2024 at 4:14 AM Josh Hunt <johunt@...mai.com> wrote:
>
> There have been multiple occassions where we have crashed in this path
> because packets_out suggested there were packets on the write or retransmit
> queues, but in fact there weren't leading to a NULL skb being dereferenced.
> While we should fix that root cause we should also just make sure the skb
> is not NULL before dereferencing it. Also add a warn once here to capture
> some information if/when the problem case is hit again.
>
> Signed-off-by: Josh Hunt <johunt@...mai.com>
> ---
>  include/net/tcp.h | 15 ++++++++++++---
>  1 file changed, 12 insertions(+), 3 deletions(-)
>
> diff --git a/include/net/tcp.h b/include/net/tcp.h
> index 2aac11e7e1cc..19ea6ed87880 100644
> --- a/include/net/tcp.h
> +++ b/include/net/tcp.h
> @@ -2433,10 +2433,19 @@ void tcp_plb_update_state_upon_rto(struct sock *sk, struct tcp_plb_state *plb);
>  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);
> +       u32 rto = jiffies_to_usecs(inet_csk(sk)->icsk_rto);
> +
> +       if (likely(skb)) {
> +               u64 rto_time_stamp_us = tcp_skb_timestamp_us(skb) + rto;
> +
> +               return rto_time_stamp_us - tcp_sk(sk)->tcp_mstamp;
> +       } else {
> +               WARN_ONCE(1,
> +                       "rtx queue emtpy: inflight %u tlp_high_seq %u state %u\n",
> +                       tcp_sk(sk)->packets_out, tcp_sk(sk)->tlp_high_seq, sk->sk_state);
> +               return rto;
> +       }
>
> -       return rto_time_stamp_us - tcp_sk(sk)->tcp_mstamp;
>  }
>
>  /*
> --
> 2.34.1
>

Are you using a recent linux kernel version ?

I am asking because sometimes patches are submitted while the authors
are using very old kernels,
and they do not state this clearly.

I have never seen such a state.

Please CC Neal on your next submission.