[Cryptech Tech] goals / use cases

[Rob Austein]

[Apologies in advance if this has been overtaken by events, catching
 up on several weeks of Cryptech mail]

At Sun, 25 Jan 2015 11:02:49 +0900, Randy Bush wrote:
...
> otoh, code complexity, threads, ... may not be what we want in the
> aplha.  i would really love to hear if paul and rob are prepared to take
> advantage multi-core in 2015.  my impression is that they are thinking
> an extremely minimal embedded world.  peter has already weighed in favor
> of an rtos.  and joachim has been inclined to the simpler M series.  so
> i am hoping for serious push-back from our various factions.

My preference would be to avoid threading unless we really need it.

My main reason for avoiding threading if possible is that any kind of
threading significantly complicates debugging and testing.  The closer
we can get to deterministic code, the easier it will be to convince
ourselves that we understand what it's doing.

That said, Peter has mentioned a couple of cases where we might decide
we need threads.  If we need them, we need them, I just don't want to
drag them in unnecessarily.  If we do need them, I would prefer that
we use them sparingly, for well-defined, well-separated tasks.

Sadly, threading has enough pitfalls (race conditions, deadlocks,
priority inversions, resource contention, ...) that multiple CPUs can
be a bit like the joke about multi-engine aircraft: the second CPU
provides just enough power to get you to the scene of the crash.

NB: Speaking only for myself here.  Paul may have his own opinion.