[tcpm] Re: Additional Security Algorithms For Use With TCP-AO

Michael Tuexen <michael.tuexen@lurchi.franken.de> Tue, 04 August 2026 17:17 UTC

Return-Path: <michael.tuexen@lurchi.franken.de>
X-Original-To: tcpm@mail2.ietf.org
Delivered-To: tcpm@mail2.ietf.org
Received: from localhost (localhost [127.0.0.1]) by mail2.ietf.org (Postfix) with ESMTP id 13D8112387267 for <tcpm@mail2.ietf.org>; Tue, 4 Aug 2026 10:17:07 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=ietf.org; s=ietf1; t=1785863827; bh=GeDaFY7GWhZ+u+be/NVtEW5s9gLS9XjMREKEtG5EvCI=; h=Subject:From:In-Reply-To:Date:Cc:References:To; b=a3tn3bP+7ftUpNaukma5dXn1xZiBsQDbZKOPc+nvQupMAakDcxIdQuvn3MuLISz6E Nx7ZA7O2WMHKjHwgfmxd/sZ6Hwmn0LsRUgIMCTndMF8cjAx+kiFm9BvK4yxZBbQrmV C7ls+azSQVLCJKq73eDFkn/axGnyWyvSIY3KQQPY=
X-Virus-Scanned: amavisd-new at ietf.org
X-Spam-Flag: NO
X-Spam-Score: -2.599
X-Spam-Level:
X-Spam-Status: No, score=-2.599 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_LOW=-0.7, RCVD_IN_VALIDITY_CERTIFIED_BLOCKED=0.001, RCVD_IN_VALIDITY_RPBL_BLOCKED=0.001, SPF_PASS=-0.001] autolearn=ham autolearn_force=no
Received: from mail2.ietf.org ([166.84.6.31]) by localhost (mail2.ietf.org [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 3wkXLYIliO4A for <tcpm@mail2.ietf.org>; Tue, 4 Aug 2026 10:17:05 -0700 (PDT)
Received: from drew.franken.de (mail-n.franken.de [193.175.24.27]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mail2.ietf.org (Postfix) with ESMTPS id 9BE7C1238725B for <tcpm@ietf.org>; Tue, 4 Aug 2026 10:17:05 -0700 (PDT)
Received: from smtpclient.apple (unknown [IPv6:2a02:8109:1101:be00:f0cd:3fc0:3cb2:76e5]) (Authenticated sender: lurchi) by drew.franken.de (Postfix) with ESMTPSA id 18CCA721BE010; Tue, 4 Aug 2026 19:16:47 +0200 (CEST)
Content-Type: text/plain; charset="utf-8"
Mime-Version: 1.0 (Mac OS X Mail 16.0 \(3864.700.51.1.1\))
From: Michael Tuexen <michael.tuexen@lurchi.franken.de>
In-Reply-To: <CAM4esxQr76Gz5RcfZjKdhLhv6Se14Xmvnc9aBvs5n=Pat6ZCkQ@mail.gmail.com>
Date: Tue, 04 Aug 2026 19:16:46 +0200
Content-Transfer-Encoding: quoted-printable
Message-Id: <1A5F48E4-42D1-4C1A-9EBD-CE01E31159B9@lurchi.franken.de>
References: <DM4PR84MB2310973EEF309B33F78CBC76F47DA@DM4PR84MB2310.NAMPRD84.PROD.OUTLOOK.COM> <85EE361A-816F-433E-96BA-7AC3134B1254@strayalpha.com> <CAM4esxQr76Gz5RcfZjKdhLhv6Se14Xmvnc9aBvs5n=Pat6ZCkQ@mail.gmail.com>
To: Martin Duke <martin.h.duke@gmail.com>
X-Mailer: Apple Mail (2.3864.700.51.1.1)
Message-ID-Hash: EA5XCF5XSHED54WBVQR4Y4GWTR2IHV4T
X-Message-ID-Hash: EA5XCF5XSHED54WBVQR4Y4GWTR2IHV4T
X-MailFrom: michael.tuexen@lurchi.franken.de
X-Mailman-Rule-Misses: dmarc-mitigation; no-senders; approved; emergency; loop; banned-address; member-moderation; header-match-tcpm.ietf.org-0; nonmember-moderation; administrivia; implicit-dest; max-recipients; max-size; news-moderation; no-subject; digests; suspicious-header
CC: Joe Touch <touch@strayalpha.com>, Ron Bonica <ronald.bonica@hpe.com>, "tcpm@ietf.org Extensions" <tcpm@ietf.org>
X-Mailman-Version: 3.3.9rc6
Precedence: list
Subject: [tcpm] Re: Additional Security Algorithms For Use With TCP-AO
List-Id: TCP Maintenance and Minor Extensions Working Group <tcpm.ietf.org>
Archived-At: <https://mailarchive.ietf.org/arch/msg/tcpm/iTU-7iF5eeRxqXNQiWDtUe3I9aw>
List-Archive: <https://mailarchive.ietf.org/arch/browse/tcpm>
List-Help: <mailto:tcpm-request@ietf.org?subject=help>
List-Owner: <mailto:tcpm-owner@ietf.org>
List-Post: <mailto:tcpm@ietf.org>
List-Subscribe: <mailto:tcpm-join@ietf.org>
List-Unsubscribe: <mailto:tcpm-leave@ietf.org>

> On 4. Aug 2026, at 18:50, Martin Duke <martin.h.duke@gmail.com> wrote:
> 
> I don't object to the efforts to extend the option space.
> 
> However, if these proposals are fraught, a potential band-aid would be a new TCP option that essentially compresses a set of TCP options that are common to TCP-AO use cases into a single option of relatively small size. The presence of this option (potentially with a few fields, as necessary) would be equivalent to a whole set of existing options.
Hi Martin,

I am not a security expert, but I don't think you can compress the output of
a MAC algorithm.

Let us have a look at the options used after the handshake:
1. TCP AO: I think this can't be compressed.
2. Time stamps: Not sure.
3. SACK blocks: This can be compressed.

However, TCP AO already fills almost the complete the option space, if
256 bit MACs are used. This corresponds to 32 bytes, which results in
an option of 36 bytes length. We might save a byte or two, but I don't
think this is enough for Timestamps or more than one SACK block.

If you have an idea how to do it, I would be very interested.

Best regards
Michael (as an individual)
> 
> On Thu, Mar 5, 2026 at 9:57 AM Joe Touch <touch@strayalpha.com> wrote:
> I agree with the first part but we have plenty of lead time while updating TCP for more option space to push the bigger hashes through then. If we list hashes that don’t currently fit, we can’t push the doc thru as standards track. 
> —
> Joe Touch, temporal epistemologist
> www.strayalpha.com
> 
>> On Mar 5, 2026, at 9:53 AM, Bonica, Ron <ronald.bonica@hpe.com> wrote:
>> 
>>  Joe,
>> 
>> You have a point. We should deploy the following algorithms as soon as possible:
>> 
>>     • HMAC-SHA224 -128
>>     • HMAC-SHA256-128 
>>     • HMAC-SHA3-224-128
>>     • HMAC-SHA3-256-128
>> 
>> But this isn't the end of the story. Authentication algorithms have a finite lifetime, and that lifetime is inversely proportional to the compute capabilities of the attacker.
>> 
>> What will we when those algorithms aren't good enough?
>> 
>> Given how long it takes to move a draft through the standardization process, we might as well progress draft-bonica-tcmp-tcp-ao-algs as is, with the caveat that many of its algorithms cannot be deployed until the 40-byte TCP Option limit is overcome.
>> 
>>                                                                                                                      Ron
>> 
>> 
>> 
>> 
>> 
>> 
>> From: touch@strayalpha.com <touch@strayalpha.com>
>> Sent: Tuesday, March 3, 2026 10:14 PM
>> To: Bonica, Ron <ronald.bonica@hpe.com>
>> Cc: Lars Eggert <lars@eggert.org>; tcpm@ietf.org Extensions <tcpm@ietf.org>
>> Subject: Re: [tcpm] Additional Security Algorithms For Use With TCP-AO 
>> 
>> Wouldn’t the SHA256 and SHA3-256 algs also qualify? Their 128-bit hash isn’t less than half the size of the key.
>> 
>> And if that’s the case, perhaps we can move forward now with just the 224 and 256 keys with the 128-bit hashes as recommended and hold on the others.
>> 
>> Joe
>> 
>> —
>> Dr. Joe Touch, temporal epistemologist
>> www.strayalpha.com
>> 
>> On Mar 3, 2026, at 10:22 AM, Bonica, Ron <ronald.bonica=40hpe.com@dmarc.ietf.org> wrote:
>> 
>> Lars,
>> 
>> Draft-bonica-tcmp-tcp-ao-algs explains how to used the following with TCP-AO:
>> 
>>     • HMAC-SHA224 
>>     • HMAC-SHA256 
>>     • HMAC-SHA384  
>>     • HMAC-SHA512 
>>     • HMAC-SHA3-224  
>>     • HMAC-SHA3-256  
>>     • HMAC-SHA3-384 
>>     • HMAC-SHA3-512  
>> 
>> All of these generate long MACs.
>> 
>> Draft-bonica-tcmp-tcp-ao-algs also explains how to used the following with TCP-AO:
>> 
>>     • HMAC-SHA224-128
>>     • HMAC-SHA256-128 
>>     • HMAC-SHA384-128  
>>     • HMAC-SHA512-128 
>>     • HMAC-SHA3-224-128  
>>     • HMAC-SHA3-256-128  
>>     • HMAC-SHA3-384-128 
>>     • HMAC-SHA3-512-128
>> 
>> All of these generate 128-bit MACs.
>> 
>> While these may serve as expedients, they are not as strong as their untruncated versions. To make matters worse, all of them except for  HMAC-SHA224-128 and HMAC-SHA3-224-128 violate the following guidance from RFC 2104:
>> 
>> "We recommend that the output length t be not less than half the length of the hash output (to match the birthday attack bound) and not less than 80 bit (a suitable lower bound on the number of bits that need to be predicted by an attacker)."
>> 
>>                                                    Ron
>> 
>> 
>> 
>> 
>> From: Lars Eggert <lars=40eggert.org@dmarc.ietf.org>
>> Sent: Tuesday, March 3, 2026 1:44 AM
>> To: Joe Touch <touch@strayalpha.com>
>> Cc: Lars Eggert <lars=40eggert.org@dmarc.ietf.org>; tcpm@ietf.org Extensions <tcpm@ietf.org>
>> Subject: [tcpm] Re: Additional Security Algorithms For Use With TCP-AO
>>   Hi,
>> 
>> On Feb 27, 2026, at 17:45, touch@strayalpha.com wrote:
>> > Modernization typically requires not only different algorithms (in some cases), but also typically longer keys and HMACs. The latter is why the option space issue keeps coming up.
>> 
>> understood. I was hoping there were *some* modern ciphersuites that we could simply squeeze into the available space now, rather than trying to thread the "extended option space" needle.
>> 
>> Thanks,
>> Lars
>> 
>> _______________________________________________
>> tcpm mailing list -- tcpm@ietf.org
>> To unsubscribe send an email to tcpm-leave@ietf.org
>> _______________________________________________
>> tcpm mailing list -- tcpm@ietf.org
>> To unsubscribe send an email to tcpm-leave@ietf.org
>> 
> _______________________________________________
> tcpm mailing list -- tcpm@ietf.org
> To unsubscribe send an email to tcpm-leave@ietf.org
> _______________________________________________
> tcpm mailing list -- tcpm@ietf.org
> To unsubscribe send an email to tcpm-leave@ietf.org