netdev - Re: [PATCH] tcp: del skb from tsorted_sent

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: <CAK6E8=eNe8Ce9zKXx1rKBL48XuDVGntAOOtKVi6ywgMjafMWXg@mail.gmail.com>
Date:   Tue, 30 Aug 2022 22:58:33 -0700
From:   Yuchung Cheng <ycheng@...gle.com>
To:     Yonglong Li <liyonglong@...natelecom.cn>
Cc:     netdev@...r.kernel.org, davem@...emloft.net, dsahern@...nel.org,
        edumazet@...gle.com, kuba@...nel.org, pabeni@...hat.com,
        Neal Cardwell <ncardwell@...gle.com>
Subject: Re: [PATCH] tcp: del skb from tsorted_sent_queue after mark it as lost

On Mon, Aug 29, 2022 at 5:23 PM Yuchung Cheng <ycheng@...gle.com> wrote:
>
> On Mon, Aug 29, 2022 at 1:21 AM Yonglong Li <liyonglong@...natelecom.cn> wrote:
> >
> > if rack is enabled, when skb marked as lost we can remove it from
> > tsorted_sent_queue. It will reduces the iterations on tsorted_sent_queue
> > in tcp_rack_detect_loss
>
> Did you test the case where an skb is marked lost again after
> retransmission? I can't quite remember the reason I avoided this
> optimization. let me run some test and get back to you.
As I suspected, this patch fails to pass our packet drill tests.

It breaks detecting retransmitted packets that
get lost again, b/c they have already been removed from the tsorted
list when they get lost the first time.



>
>
> >
> > Signed-off-by: Yonglong Li <liyonglong@...natelecom.cn>
> > ---
> >  net/ipv4/tcp_input.c    | 15 +++++++++------
> >  net/ipv4/tcp_recovery.c |  1 -
> >  2 files changed, 9 insertions(+), 7 deletions(-)
> >
> > diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
> > index ab5f0ea..01bd644 100644
> > --- a/net/ipv4/tcp_input.c
> > +++ b/net/ipv4/tcp_input.c
> > @@ -1082,6 +1082,12 @@ static void tcp_notify_skb_loss_event(struct tcp_sock *tp, const struct sk_buff
> >         tp->lost += tcp_skb_pcount(skb);
> >  }
> >
> > +static bool tcp_is_rack(const struct sock *sk)
> > +{
> > +       return READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_recovery) &
> > +               TCP_RACK_LOSS_DETECTION;
> > +}
> > +
> >  void tcp_mark_skb_lost(struct sock *sk, struct sk_buff *skb)
> >  {
> >         __u8 sacked = TCP_SKB_CB(skb)->sacked;
> > @@ -1105,6 +1111,9 @@ void tcp_mark_skb_lost(struct sock *sk, struct sk_buff *skb)
> >                 TCP_SKB_CB(skb)->sacked |= TCPCB_LOST;
> >                 tcp_notify_skb_loss_event(tp, skb);
> >         }
> > +
> > +       if (tcp_is_rack(sk))
> > +               list_del_init(&skb->tcp_tsorted_anchor);
> >  }
> >
> >  /* Updates the delivered and delivered_ce counts */
> > @@ -2093,12 +2102,6 @@ static inline void tcp_init_undo(struct tcp_sock *tp)
> >         tp->undo_retrans = tp->retrans_out ? : -1;
> >  }
> >
> > -static bool tcp_is_rack(const struct sock *sk)
> > -{
> > -       return READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_recovery) &
> > -               TCP_RACK_LOSS_DETECTION;
> > -}
> > -
> >  /* If we detect SACK reneging, forget all SACK information
> >   * and reset tags completely, otherwise preserve SACKs. If receiver
> >   * dropped its ofo queue, we will know this due to reneging detection.
> > diff --git a/net/ipv4/tcp_recovery.c b/net/ipv4/tcp_recovery.c
> > index 50abaa9..ba52ec9e 100644
> > --- a/net/ipv4/tcp_recovery.c
> > +++ b/net/ipv4/tcp_recovery.c
> > @@ -84,7 +84,6 @@ static void tcp_rack_detect_loss(struct sock *sk, u32 *reo_timeout)
> >                 remaining = tcp_rack_skb_timeout(tp, skb, reo_wnd);
> >                 if (remaining <= 0) {
> >                         tcp_mark_skb_lost(sk, skb);
> > -                       list_del_init(&skb->tcp_tsorted_anchor);
> >                 } else {
> >                         /* Record maximum wait time */
> >                         *reo_timeout = max_t(u32, *reo_timeout, remaining);
> > --
> > 1.8.3.1
> >