Please please read the User Guide. Nothing will make sense otherwise (it still may not afterwards, but at least you've a better chance).
No new validations are currently planned.
FIPS What? Where Do I Start?
Ok, so your company needs FIPS validated cryptography to land that big sale, and your product currently uses OpenSSL. You haven't worked up the motivation to wade through the entire User Guide and want the quick "executive summary". Here is a grossly oversimplified account:
- OpenSSL itself is not validated, and never will be. Instead a special carefully defined software component called the OpenSSL FIPS Object Module has been created. This Module was designed for compatibility with OpenSSL so that products using the OpenSSL API can be converted to use validated cryptography with minimal effort.
- The OpenSSL FIPS Object Module validation is unique among all FIPS 140-2 validations in that the product is "delivered" in source code form, meaning that if you can use it exactly as is and can build it (according to the very specific documented instructions) for your platform, then you can use it as validated cryptography on a "vendor affirmed" basis.
- If even the tiniest source code or build process changes are required for your intended application, you cannot use the open source based validated module directly. You must obtain your own validation. This situation is common; see "Private Label" validation, below.
- New FIPS 140-2 validations (of any type) are slow (6-12 months is typical), expensive (US$50,000 is probably typical for an uncomplicated validation), and unpredictable (completion dates are not only uncertain when first beginning a validation, but remain so during the process).
Note that FIPS 140-2 validation is a complicated topic that the above summary does not adequately address. You have been warned!
We refer to validations based directly on the OpenSSL FIPS Object Module as "private label" validations. These are also sometimes referred to as "cookie cutter" validations. The usual reason for such separate validations is the need for small modifications which forces a complete new validation, but some vendors, for marketing or risk management reasons, have obtained private label validations for binaries produced from unmodified (or only cosmetically modified) source code.
The OSF would really prefer to work on open source based validations of benefit to the OpenSSL user community at large, but financial support for that objective is intermittent at best. On the other hand many vendors are interested in private label validations and the OSF will assist in such efforts on a paid basis. We've done enough of these to be very cost competitive, and for uncomplicated validations we typically work on a fixed price basis.
Update: As of 2015 we are no longer performing private label validations. We are still adding new platforms to the #1747 or related validations.
Performance at Startup
We have had many complaints about poor performance of the Power-On Self Test (POST) on low powered computers, as with some embedded devices. In the worst cases the POST can take several minutes. Such devices were not included as test platforms at the time the code was originally written.
The current FIPS validated code performs a very comprehensive set of mandatory algorithm self tests when it enter FIPS mode covering many algorithm combinations. There is a DSA parameter generation self test which is especially CPU intensive.
As a result of the POST performance issue we revisited the KAT (Known Answer Test) requirements in the POST process that were burning up most of those cycle. In consultation with a CMVP test lab we determined that it should be possible to substantially reduce that performance penalty in a new validation. Unfortunately such a change can only be undertaken in the context of a new validation, and not as a change letter modification.
Another factor affecting performance is the use (or not) of platform specific optimizations. The x86/x64 Windows and Linux code makes use of assembly language optimizations for FIPS cryptographic algorithms. The C only version is much slower and so the POST is slower too.