linux-kernel - Re: [PATCH net-next] tcp: clamp window like before the cleanup

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] [day] [month] [year] [list]

Message-ID: <58d604b3-760f-449d-95b4-c56d8b9da400@kernel.org>
Date: Thu, 6 Mar 2025 12:20:29 +0100
From: Matthieu Baerts <matttbe@...nel.org>
To: Eric Dumazet <edumazet@...gle.com>
Cc: Jason Xing <kerneljasonxing@...il.com>, mptcp@...ts.linux.dev,
 Neal Cardwell <ncardwell@...gle.com>, Kuniyuki Iwashima <kuniyu@...zon.com>,
 "David S. Miller" <davem@...emloft.net>, David Ahern <dsahern@...nel.org>,
 Jakub Kicinski <kuba@...nel.org>, Paolo Abeni <pabeni@...hat.com>,
 Simon Horman <horms@...nel.org>, netdev@...r.kernel.org,
 linux-kernel@...r.kernel.org
Subject: Re: [PATCH net-next] tcp: clamp window like before the cleanup

Hi Eric,

On 06/03/2025 12:08, Eric Dumazet wrote:
> On Thu, Mar 6, 2025 at 11:16 AM Eric Dumazet <edumazet@...gle.com> wrote:
>> On Thu, Mar 6, 2025 at 11:12 AM Matthieu Baerts <matttbe@...nel.org> wrote:
>>> On 06/03/2025 11:02, Eric Dumazet wrote:

(...)

>>>> I am wondering if this could hide an issue in MPTCP ?
>>> Indeed, I was wondering the same thing. I didn't see anything obvious
>>> when looking at this issue. The behaviours around the window clamping,
>>> with MPTCP single flow, and "plain" TCP were quite similar I think.
>>
>> OK, let me run mptcp tests just in case I see something dubious.
> 
> I have no idea why only MPTCP flows can trigger the condition, I do
> not think it matters anyway.

Thank you for having looked! Did you manage to trigger this condition
with "plain" TCP (without MPTCP) when using "./mptcp_connect.sh -tt"?

The TCP only connections will be printed like this:

  nsX TCP   -> nsY (<IP>:<port>) TCP

But yes, it probably doesn't matter.

Cheers,
Matt
-- 
Sponsored by the NGI0 Core fund.