[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CADVnQynUxv-+7HBUR_60jE-nKv5ix+e51L8N4=px1JUeMuqYhg@mail.gmail.com>
Date: Mon, 27 Jun 2016 11:47:45 -0400
From: Neal Cardwell <ncardwell@...gle.com>
To: Eric Dumazet <eric.dumazet@...il.com>
Cc: Arjun V <arjun@...lsio.com>, Eric Dumazet <edumazet@...gle.com>,
"netdev@...r.kernel.org" <netdev@...r.kernel.org>,
Hariprasad S <hariprasad@...lsio.com>,
Casey Leedom <leedom@...lsio.com>,
Kumar A S <kumaras@...lsio.com>,
Santosh Rastapur <santosh@...lsio.com>,
Nirranjan Kirubaharan <nirranjan@...lsio.com>,
"davem@...emloft.net" <davem@...emloft.net>,
"ycheng@...gle.com" <ycheng@...gle.com>
Subject: Re: [PATCH net] tcp: do not send too big packets at retransmit time
On Mon, Jun 27, 2016 at 11:38 AM, Eric Dumazet <eric.dumazet@...il.com> wrote:
> From: Eric Dumazet <edumazet@...gle.com>
>
> Arjun reported a bug in TCP stack and bisected it to a recent commit.
>
> In case where we process SACK, we can coalesce multiple skbs
> into fat ones (tcp_shift_skb_data()), to lower write queue
> overhead, because we do not expect to retransmit these packets.
>
> However, SACK reneging can happen, forcing the sender to retransmit
> all these packets. If skb->len is above 64KB, we then send buggy
> IP packets that could hang TSO engine on cxgb4.
>
> Neal suggested to use tcp_tso_autosize() instead of tp->gso_segs
> so that we cook packets of optimal size vs TCP/pacing.
>
> Thanks to Arjun for reporting the bug and running the tests !
>
> Fixes: 10d3be569243 ("tcp-tso: do not split TSO packets at retransmit time")
> Signed-off-by: Eric Dumazet <edumazet@...gle.com>
> Reported-by: Arjun V <arjun@...lsio.com>
> Tested-by: Arjun V <arjun@...lsio.com>
> ---
> net/ipv4/tcp_output.c | 7 ++++++-
> 1 file changed, 6 insertions(+), 1 deletion(-)
Acked-by: Neal Cardwell <ncardwell@...gle.com>
Thanks, Eric!
neal
Powered by blists - more mailing lists