On Wed, Aug 21, 2013 at 5:41 AM, Ben Laurie <[email protected]> wrote: > Something needs to be done, but won't this re-introduce the problem of > /dev/random starvation, leading to more use of /dev/urandom (on platforms > where this is a problem)? > > Mixing in the time seems like a safer solution that should also fix the > problem. Possibly only when the PID changes.
Stirring in time and PID seems like just a fail-safe. Some bytes from /dev/urandom should also be added -- it won't hang once seeded (or ever on Linux, but hopefully a simple service can be added by users to seed urandom from random). Provided one read from /dev/random has been done I think perturbing the pool with time + PID + urandom should suffice. Nico -- ______________________________________________________________________ OpenSSL Project http://www.openssl.org User Support Mailing List [email protected] Automated List Manager [email protected]
