Re: [Curdle] Which curves are MUST and SHOULD ?

Ron Frederick <ronf@timeheart.net> Tue, 05 January 2021 02:40 UTC

Return-Path: <ronf@timeheart.net>
X-Original-To: curdle@ietfa.amsl.com
Delivered-To: curdle@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id E025C3A0D5D for <curdle@ietfa.amsl.com>; Mon, 4 Jan 2021 18:40:38 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.099
X-Spam-Level:
X-Spam-Status: No, score=-2.099 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, SPF_HELO_NONE=0.001, SPF_PASS=-0.001, URIBL_BLOCKED=0.001] autolearn=unavailable autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=timeheart.net
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id ahbbvN9zw0nP for <curdle@ietfa.amsl.com>; Mon, 4 Jan 2021 18:40:37 -0800 (PST)
Received: from mail-pj1-x1033.google.com (mail-pj1-x1033.google.com [IPv6:2607:f8b0:4864:20::1033]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 7E2BD3A0D6B for <curdle@ietf.org>; Mon, 4 Jan 2021 18:40:37 -0800 (PST)
Received: by mail-pj1-x1033.google.com with SMTP id l23so829172pjg.1 for <curdle@ietf.org>; Mon, 04 Jan 2021 18:40:37 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=timeheart.net; s=mail; h=mime-version:subject:from:in-reply-to:date:cc :content-transfer-encoding:message-id:references:to; bh=XUPQFufTa7JEKg2x7l1a7DyFc/ibAtmkDmRk87T8ZRI=; b=ZTCh25WEyac4Cb7byZrHZi42OHBJ621TJHdwxiuMsHPKcYoDtV7+xZqde5tTZJ0Gwj 6Uabbnim/c/nRMezlfCPWvjAxq4Jok5Z/z2Gnyj1vk6A4CCNH/NpPYFpVbvqTehoCgpL VTt+bHGqh4MC75HwTiEbaz6ycvI4rnj/V4ErA=
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:subject:from:in-reply-to:date:cc :content-transfer-encoding:message-id:references:to; bh=XUPQFufTa7JEKg2x7l1a7DyFc/ibAtmkDmRk87T8ZRI=; b=FNsdJWsvIU+Pjeiexe7vrjtaAsUSHmnjTlyNAiJfFD9vgBGgCpeTrffyWtGC1zv97z bWP/2KNgr1XNrlFis5uBmflZVZ70V1mOHlFC89Clg6zyBpsxtXwdGMZAi3SEF6RNoiSk qtXMkYOjLMjEQRu6KcfEWtLAmo23xc4p/Wfu0g2lIuSAAiGh17qY0tSzK9077lDtP8rv Fj72KwZtuFTFeDqEXXPqeHcHnmKkn+r/t0mZ6yzyYLSOPD1O3bAQhOg4lFuNypMfLTsq H1f84yEF3q+Go1xNUsH7RMzaiH4NG+234VJUzeF1oNFlV5/epFXsqiZ5nClXQ74In4zj VxcA==
X-Gm-Message-State: AOAM530C85NUzlylpxGGYF4H3Kz/hmGGxpqDgMoi0ag1oSYtZwA8TFD+ fbaFRlYUtUnAuB85y0ol7nXCwg==
X-Google-Smtp-Source: ABdhPJzMFxFGDQgNuQCt4d4HLIk7drBOTuyiKE54NI9ihUXbjJzgpOvBW1Kyn6LSCOAudIL/1OrTcg==
X-Received: by 2002:a17:90a:c085:: with SMTP id o5mr1904528pjs.210.1609814435868; Mon, 04 Jan 2021 18:40:35 -0800 (PST)
Received: from connect.timeheart.net.0.4.a.f.8.1.4.2.0.3.3.0.6.2.ip6.arpa ([2603:3024:18fa:4000::2]) by smtp.gmail.com with ESMTPSA id 17sm30706994pgy.53.2021.01.04.18.40.35 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Mon, 04 Jan 2021 18:40:35 -0800 (PST)
Content-Type: text/plain; charset=utf-8
Mime-Version: 1.0 (Mac OS X Mail 12.4 \(3445.104.14\))
From: Ron Frederick <ronf@timeheart.net>
In-Reply-To: <90135.1609791710@eng-mail03>
Date: Mon, 4 Jan 2021 18:40:34 -0800
Cc: Hubert Kario <hkario@redhat.com>, curdle@ietf.org
Content-Transfer-Encoding: quoted-printable
Message-Id: <BB60E7A2-B3EF-4DF0-A346-821D14D632DE@timeheart.net>
References: <2CCABC30-F757-4659-9FF3-5AADDD51EE30@akamai.com> <4b681efd49274f03c7e0521e127e031426632ad0.camel@redhat.com> <CADZyTkk--kCWqE7q0Xi5C40V92MuZBktDzQGt_vPSZPiBy7v9w@mail.gmail.com> <18479.1606885358@eng-mail01.juniper.net> <20201205194724.GB64351@kduck.mit.edu> <37691.1607621661@eng-mail01.juniper.net> <1607647129866.76532@cs.auckland.ac.nz> <2917.1607672034@eng-mail01.juniper.net> <012AE120-2516-44F6-B729-ED342A137535@timeheart.net> <ED8F3B46-A5CC-4D14-A714-FD1C0AA67486@akamai.com> <12959BD6-F3AB-418B-8CE0-C3BE43999435@timeheart.net> <40887.1608233724@eng-mail03> <0f4dce32-b362-43d8-85e0-9608ca3427ab@redhat.com> <90135.1609791710@eng-mail03>
To: "Mark D. Baushke" <mdb=40juniper.net@dmarc.ietf.org>
X-Mailer: Apple Mail (2.3445.104.14)
Archived-At: <https://mailarchive.ietf.org/arch/msg/curdle/JMqowtaWCwOfuMBQlxDz45r3wQs>
Subject: Re: [Curdle] Which curves are MUST and SHOULD ?
X-BeenThere: curdle@ietf.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: "List for discussion of potential new security area wg." <curdle.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/curdle>, <mailto:curdle-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/curdle/>
List-Post: <mailto:curdle@ietf.org>
List-Help: <mailto:curdle-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/curdle>, <mailto:curdle-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 05 Jan 2021 02:40:39 -0000

On Jan 4, 2021, at 12:21 PM, Mark D. Baushke <mdb=40juniper.net@dmarc.ietf.org> wrote:
> Two kinds of stakeholders: a) "implementors" and b) "users" should mean
> more responses for the question.
> 
> Okay. In the original RFC4253 specification both
> 
>    diffie-hellman-group1-sha1 
> and
>    diffie-hellman-group14-sha1
> 
> were REQUIRED key exchanges.
> 
> The group1 parameters in RFC4253 point to the 1024-bit MODP Second
> Oakley Group given in RFC2409 section 6.2 and RFC2412 section E.2.
> 
> There are two issues with diffie-hellman-group1-sha1: 1) recent
> estimages are that it has roughly 80 bits of security strength, and 2)
> it uses SHA1 for hashing which is considered weak.
> 
> If we choose "MUST NOT" for this key exchange, then we are going from
> "MUST" to "MUST NOT" which could be a hardship for low-end devices
> unable to run calculations to generate a shared secret using a larger
> MODP group if support is completely removed.
> 
> If we choose "SHOULD NOT", then it is hoped that most implementors would
> default to not configuring this option by default, but may provide it
> for enviornments that need it.
> 
> If we choose "MAY", then it is not certain if implementors or users will
> do much of anything different and this potentially insecure key exchange
> may continue to be used even when it may be a hazard to those that
> desire a more secure by default system.
> 
> Are you an SSH impelmentor or user or both?
> 
>  Implementor
>  User
>  Both

[Ron] Primarily implementor for this discussion, but I’m also definitely an SSH end-user (for both client and server).


> I would like to get a straw vote for the six *sha1* related key
> exchanges. I am proposing that the rsa1024-sha1-* kex be a MUST NOT and
> that all of the others be a SHOULD NOT.
> 
> 1. For diffie-hellman-group1-sha1 what is your vote?
> 
>  MUST          -- current for RFC4253
>  SHOULD
>  MAY
>  SHOULD NOT    -- proposed in the -13 draft
>  MUST NOT

[Ron] I’m good with SHOULD NOT here, and would take this to mean that this could still be explicitly enabled but would be disabled by default in my implementation.


> 2. For diffie-hellman-group14-sha1 what is your vote?
> 
>  MUST          -- current for RFC4253
>  SHOULD
>  MAY           -- proposed in the -13 draft
>  SHOULD NOT
>  MUST NOT

[Ron] I’m good with MAY here, and would consider group14-sha1 as the preference to maintain interoperability with older/slower devices.


> 3. For diffie-hellman-group-exchange-sha1 what is your vote?
> 
>  MUST
>  SHOULD
>  MAY           -- current for RFC4419    
>  SHOULD NOT    -- proposed in the -13 draft
>  MUST NOT

[Ron] I’m good with SHOULD NOT here.


> 4. For rsa1024-sha1 what is your vote?
> 
>  MUST
>  SHOULD
>  MAY           -- current for RFC4432
>  SHOULD NOT
>  MUST NOT      -- proposed in the -13 draft

[Ron] I’d be ok with either SHOULD NOT or MUST NOT on this one. I’d lean toward SHOULD NOT to be consistent with other sha1 cases here and to give a bit more time for people to move off of rsa1024-sha1 before implementations start removing it completely. However, I’m not sure it was ever that popular to begin with, so I don’t really know how necessary it is to do this in two phases.


> 5. For gss-gex-sha1-* what is your vote?
> 
>  MUST
>  SHOULD        -- current for RFC4462
>  MAY
>  SHOULD NOT    -- proposed in the -13 draft
>  MUST NOT  
> 
> 6. For gss-group1-sha1-* what is your vote?
> 
>  MUST
>  SHOULD        -- current for RFC4462
>  MAY
>  SHOULD NOT    -- proposed in the -13 draft
>  MUST NOT  


[Ron] I’m good with SHOULD NOT for both of these.
-- 
Ron Frederick
ronf@timeheart.net