[Info-vax] In need of "good enough" randomness sources, anyone wanna help out? (OpenSSL 1.1.1 beta)

[Stephen Hoffman]

On 2018-05-01 07:10:19 +0000, Richard Levitte said:

> So I did a bit of experiment and collected more data, most of all quite 
> a lot of the RMI$_ items but also a few more JPI$_ items, at a total of 
> 736 bytes per gathering round and a "random" 1 to 10 second sleep 
> between rounds, fed that into a file that I then fed NISTs minimum 
> entropy estimation program (*)...  and the result was pretty harsh, an 
> estimate of 0.082 entropy bits per 8 bits of data (**), so roughly 1 
> bit of entropy per 100 bits of data.

I'd liberally mix in $purge_ws (for 0x0-0xffffffffffffffffff) and 
$resched calls and maybe toss around some ASTs using some mix of those 
into your existing code trying to churn up some entropy, and through 
all that sprinkle multiple clock_gettime (V7.3-2 and later) or 
sys$rpcc_64 calls or rpcc built-in calls or probably better rscc 
built-in calls, all in the service of added entropy.  Maybe add a few 
briefly-present busy-work compute-bound subprocesses while all that's 
running; invoking the subprocesses using some busy-work option embedded 
in the main OpenSSL utility or otherwise.  The Alpha rpcc and rscc 
stuff has been around for a long while, and the APIs did get ported to 
Itanium.  http://h41379.www4.hpe.com/wizard/wiz_4004.html et al.  
This'd all be far easier if, you know, there was an entropy pool in the 
kernel and APIs, but that's fodder for some future OpenVMS release.


-- 
Pure Personal Opinion | HoffmanLabs LLC