[email protected] writes:

> On Mon, Dec 27, 2010 at 6:47 AM, Henrik Grindal Bakken <[email protected]> 
> wrote:
>>> 3) obtaining your own from-scratch validation starting with the
>>> 1.0.0 baseline.  Good luck with that, you have a long row to hoe.
>>
>> We're going for 3), but as I said, our crypto module is not OpenSSL,
>> it's the entire product, so the OpenSSL FIPS Object Module isn't
>> interesting for me.  What is interesting, however, are the self-tests.
>>
>>> Incidentally, unless you're seeking a Level 2 validation for a
>>> non-CC certified environment you'll regret defining the crypto
>>> module boundary to include your entire application.
>>
>> I am seeking a level 2 validation.  It's not really an application,
>> it's a hardware device.
>
> Because the 1.0.x releases don't have any support for FIPS, they
> don't implement the FIPS-mandated tests.  OpenSSL can't help with
> prevalidation now, since its FIPS-validated mode fails new
> validation standards compliance in some manner.
>
> You're probably already dealing with having to fork the code (to
> prevent any use of non-FIPS-approved algorithms when your hardware
> is operating in FIPS-validated mode).  This suggests that your best
> option might be to import the self-tests from fips-1.2.2 into your
> own fork.

It's not really a fork; we do it mostly with a) configuration, and b)
limiting algorithms in our own code.  We do have some patches, though,
and I suppose the best way forward is to pull in the tests from the
latest OpenSSL FIPS module and try to fit them into 1.0.0 code.

Thanks for your advice.


-- 
Henrik Grindal Bakken <[email protected]>
PGP ID: 8D436E52
Fingerprint: 131D 9590 F0CF 47EF 7963  02AF 9236 D25A 8D43 6E52
______________________________________________________________________
OpenSSL Project                                 http://www.openssl.org
User Support Mailing List                    [email protected]
Automated List Manager                           [email protected]

Reply via email to