-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

On 09/02/14 10:09, Peter Backes via RT wrote:
> SHA512 is not offered at all

To the best of my knowlege, there is no ciphersuite that uses
SHA-512, either in OpenSSL, or in the RFCs. OpenSSL cannot
implement a ciphersuite until identifiers for it have been
assigned, it would hamper interoperability with other
cryptographic toolkits.

> for AES256; SHA384 is the maximum.

This is in line with the NSA Suite B specification. Also,
SHA-384 is the same algorithm as SHA-512, just with different
starting constants, and truncation to reduce its output length.

> Further, SHA1 appears in quite a lot of ciphers suites. It has
> severe known weaknesses.

No it does not. To the best of my knowlege, no-one has
demonstrated a feasible second-preimage attack on SHA-1.

However, its use is starting to be discouraged now, for
example in NIST SP800-131A, when used in a signature scheme
by itself. For example, SHA1withRSA in X.509 certificates, or
SHA1 in PGP (you will note I am PGP-signing this message with
SHA256).

But more to the point, SHA-1 as used in ciphersuites is used
in an HMAC construction. Deficiencies or vulnerabilities that
affect plain SHA-1 do not affect HMAC-SHA-1. For this reason,
even HMAC-MD5 is still *okay* (although its use in new
applications is discouraged).

> Yet, unreasonable cipher suites with SHA1, like 
> ECDHE-ECDSA-AES128-SHA even have higher priority than reasonable
> ones like DHE-RSA-AES128-SHA256.

Applications are free to set their own order of ciphers with
SSL_CTX_set_cipher_list(3ssl). Also, servers are free to use
their preferred order of ciphersuites, rather than the first
suite from the client list that is mutually supported, by
setting the SSL_OP_CIPHER_SERVER_PREFERENCE option.

- -- 
Aaron Jones.


-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/
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=CRFH
-----END PGP SIGNATURE-----
______________________________________________________________________
OpenSSL Project                                 http://www.openssl.org
Development Mailing List                       [email protected]
Automated List Manager                           [email protected]

Reply via email to