Re: [saag] Algorithms/modes requested by users/customers

[Jack Lloyd <[e-mail]>]

On Tue, Feb 19, 2008 at 08:18:24AM -0800, Paul Hoffman wrote:

> - Some of the tests are fairly subjective, and it becomes a game of 
> fixing code to please the testing service, not to make the product 
> more secure.
[...]
> - The system introduces silly modes that make the systems more complicated.

I have no experience with the purchasing side, but in my experience
doing FIPS 140 validations, we often had to ask vendors to include
hooks for testing that, from any objective standpoint, made the system
less secure. And because the tests must be made on the same
firmware/software as the as-shipped one (not in a special test/debug
mode), that increased the attack surface of some of these devices
greatly. I will fondly remember the validation where I found several
exploitable buffer overflows in an HSM that had already passed two
previous validations - all the holes were found in the hooks used for
FIPS-140 testing.

The tests that require the RNG be able to seeded with fixed data
always seeemed particularly troublesome/dangerous to me.

Obvious disclaimer: Anecdotes are not data - I just thought a
relatively concrete example might be relevant to the discussion at
hand.

-Jack