[Cryptography] DH non-prime kills "socat" command security

[mok-kong shen]

Am 03.02.2016 um 02:43 schrieb Henry Baker:
> For the past year, the Linux command "socat" has been assuming that the
> following number is prime; thus breaking its crypto security.
[snip]

Among possible causes I surmise that possibly Miller-Rabin test
was used to find a prime, which however can't guarantee that the
result is a prime (only highly likely a prime when a sufficient
number of rounds of the test get passed).

Anyway, I suppose one should in crypto replace prime generation
with probability tests by provable prime generatión. I have coded
Maurer's algorithm of generation of provable primes in Python and
found that (at least in Python, which is interpreted and hence less
runtime efficient) Maurer's algorithm is quite comparable in cpu
time with the method that uses the Miller-Rabin test. See
http://s13.zetaboards.com/Crypto/topic/7234475/1/

M. K. Shen