Re: [TLS] Using RSA PSS in TLS (Martin Rex) Mon, 14 October 2013 12:12 UTC

Return-Path: <>
Received: from localhost (localhost []) by (Postfix) with ESMTP id A736311E817E for <>; Mon, 14 Oct 2013 05:12:44 -0700 (PDT)
X-Virus-Scanned: amavisd-new at
X-Spam-Flag: NO
X-Spam-Score: -9.899
X-Spam-Status: No, score=-9.899 tagged_above=-999 required=5 tests=[AWL=0.350, BAYES_00=-2.599, HELO_EQ_DE=0.35, RCVD_IN_DNSWL_HI=-8]
Received: from ([]) by localhost ( []) (amavisd-new, port 10024) with ESMTP id sD731TEO-v5G for <>; Mon, 14 Oct 2013 05:12:39 -0700 (PDT)
Received: from ( []) by (Postfix) with ESMTP id 34AD111E8138 for <>; Mon, 14 Oct 2013 05:12:35 -0700 (PDT)
Received: from by (26) with ESMTP id r9ECCYLi025617 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=OK); Mon, 14 Oct 2013 14:12:34 +0200 (MEST)
In-Reply-To: <>
To: Johannes Merkle <>
Date: Mon, 14 Oct 2013 14:12:34 +0200 (CEST)
X-Mailer: ELM [version 2.4ME+ PL125 (25)]
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Type: text/plain; charset="US-ASCII"
Message-Id: <>
From: (Martin Rex)
X-SAP: out
Cc: "<>" <>
Subject: Re: [TLS] Using RSA PSS in TLS
X-Mailman-Version: 2.1.12
Precedence: list
List-Id: "This is the mailing list for the Transport Layer Security working group of the IETF." <>
List-Unsubscribe: <>, <>
List-Archive: <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>
X-List-Received-Date: Mon, 14 Oct 2013 12:12:44 -0000

Johannes Merkle wrote:
> While the current discussion on this list is about ECC,
> I would like to raise the question if it were not desirable to
> allow usage of provably secure RSA-PSS signatures from PKCS#1v2.1 in TLS.
> Of course, an issue with this idea is that, if we replace DHE_RSA and
> ECDHE_RSA with DHE_PSS and ECDHE_PSS, we end up
> with a whole bunch of new cipher suites. An alternative could be a
> new extension signaling the RSA version used (with
> PKCS#1v1.5 as default).
> Any thoughts on that?

RSA-PSS would apply only to the DHE_RSA variants, for the vanilla
RSA ciphersuites, one would need RSA-OAEP instead, and it seems
that using a traditional RSA key/cert for both, signature and
encryption is somewhat incompatible with RSA-PSS/RSA-OAEP -- when
signaling is through the AlgorithmIdentifier in SPKI of the
entity certificate.

Having to support two seperate server certificate (different by algorithm)
would have a huge impact on
   - TLS implementations
   - administrative UIs
   - server administration procedures
   - CA procedures

What exactly is the real and serious problem that such a HUGE change
would be trying to fix?