netdev - Re: [PATCH net-next 1/2] tcp: add a common helper to debug the underlying issue

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: <CAL+tcoByVRMTt5r1WPN8ovwQDW0fO-ksWya-MCMw2v_93DOCLQ@mail.gmail.com>
Date: Mon, 21 Oct 2024 00:34:31 +0800
From: Jason Xing <kerneljasonxing@...il.com>
To: David Ahern <dsahern@...nel.org>
Cc: davem@...emloft.net, edumazet@...gle.com, kuba@...nel.org, 
	pabeni@...hat.com, ncardwell@...gle.com, netdev@...r.kernel.org, 
	Jason Xing <kernelxing@...cent.com>
Subject: Re: [PATCH net-next 1/2] tcp: add a common helper to debug the
 underlying issue

Hello David,

On Mon, Oct 21, 2024 at 12:18 AM David Ahern <dsahern@...nel.org> wrote:
>
> On 10/20/24 8:50 AM, Jason Xing wrote:
> > From: Jason Xing <kernelxing@...cent.com>
> >
> > Following the commit c8770db2d544 ("tcp: check skb is non-NULL
> > in tcp_rto_delta_us()"), we decided to add a helper so that it's
> > easier to get verbose warning on either cases.
> >
> > Link: https://lore.kernel.org/all/5632e043-bdba-4d75-bc7e-bf58014492fd@redhat.com/
> > Suggested-by: Paolo Abeni <pabeni@...hat.com>
> > Signed-off-by: Jason Xing <kernelxing@...cent.com>
> > Cc: Neal Cardwell <ncardwell@...gle.com>
> > ---
> >  include/net/tcp.h | 28 +++++++++++++++++-----------
> >  1 file changed, 17 insertions(+), 11 deletions(-)
> >
> > diff --git a/include/net/tcp.h b/include/net/tcp.h
> > index 739a9fb83d0c..cac7bbff61ce 100644
> > --- a/include/net/tcp.h
> > +++ b/include/net/tcp.h
> > @@ -2430,6 +2430,22 @@ void tcp_plb_update_state(const struct sock *sk, struct tcp_plb_state *plb,
> >  void tcp_plb_check_rehash(struct sock *sk, struct tcp_plb_state *plb);
> >  void tcp_plb_update_state_upon_rto(struct sock *sk, struct tcp_plb_state *plb);
> >
> > +static inline void tcp_warn_once(const struct sock *sk, bool cond, const char *str)
> > +{
> > +     WARN_ONCE(cond,
> > +               "%s"
> > +               "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",
>
> format lines should not be split across lines. Yes, I realize the
> existing code is, but since you are moving it and making changes to it
> this can be fixed as well.

Thanks for reminding me. Actually before submitting this series, I
noticed this warning. I was thinking it looks a little ugly if we are
going to add more information in this function?

This function could be like this:
diff --git a/include/net/tcp.h b/include/net/tcp.h
index 739a9fb83d0c..8b8d94bb1746 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -2430,6 +2430,19 @@ void tcp_plb_update_state(const struct sock
*sk, struct tcp_plb_state *plb,
 void tcp_plb_check_rehash(struct sock *sk, struct tcp_plb_state *plb);
 void tcp_plb_update_state_upon_rto(struct sock *sk, struct tcp_plb_state *plb);

+static inline void tcp_warn_once(const struct sock *sk, bool cond,
const char *str)
+{
+       WARN_ONCE(cond,
+                 "%sout:%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",
+                 str,
+                 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);
+}
+
That quoted line seems a little long... Do you think this format is
fine with you? If so, I will adjust it in the next version.

Thanks,
Jason