Re: [Qemu-discuss] Problem with AES 256 algorithm / GCM mode.

[Jakob Bohm]

On 10/8/2012 5:04 PM, MACH Christian wrote:
> Hello.
>
> I use OpenSSL for my work and particularly the AES 256 algorithm with
> the GCM mode.
>
> When I test this mode, the ciphered text is correct but the
> authentication tag is not correct.
>
> I think my test vectors are correct (source : NIST and my cipher room).
>
> Could you help me? If yes, how can we proceed?
1. Given where you appear to work, I think you should not discuss this
in public,most people on this list probably have less GCHQ/NSA
clearancethanyou do (I certainly have no clearance).  Some probably
work for currentwar enemies of your country (I do not).

2. Look in the actual GCM source code inside OpenSSL and (for test
purposes *only* call that code directly to determine if the
tag/key/parametersare accidentally garbled on the way between your
calls to the public interfacesand the actual code.

3. Also compare that source code with the algorithm and key mode
specifications you got from your cipher room to see if there is any
obvious discrepancy.  Then compare that aspect with the published
unclassified specifications that this OpenSSL code claims to follow.

Step 1 should tell you if there is a bug in the OpenSSL higher level
interfaces (such as EVP) or your understanding of how to call those.

Step 2 should tell you if the OpenSSL code is implementing the same
"GCM mode" needed by your requirements.

Step 3 should tell you if OpenSSL is failing to do what it claims to
door if your project requirements specify a different GCM mode than
theone published for civilian use.

Enjoy

Jakob
--
Jakob Bohm, CIO, Partner, WiseMo A/S. http://www.wisemo.com
Transformervej 29, 2730 Herlev, Denmark. Direct +45 31 13 16 10
This public discussion message is non-binding and may contain errors.
WiseMo - Remote Service Management for PCs, Phones and Embedded