More perf tests

[fjarri]

• Add performance tests for realistic Sieve iterations
• Add performance tests against glass-pumpkin

glass-pumpkin uses more MR tests than required by BPSW, so in the performance tests I used custom method to mimic that behavior. The results (against crypto-bigint=0.6.0-rc.2):

1024 bits:

• crypto-primes default: 29ms
• crypto-primes, matched with glass-pumpkin: 39ms
• glass-pumpkin: 46ms

1024 bits, safe prime:

• crypto-primes default: 1.33s
• crypto-primes, matched with glass-pumpkin: 1.4s
• glass-pumpkin: 1.43s

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 99.33%. Comparing base (08c0e5f) to head (258d153).
Report is 4 commits behind head on master.

Additional details and impacted files

@@           Coverage Diff           @@
##           master      #49   +/-   ##
=======================================
  Coverage   99.33%   99.33%           
=======================================
  Files           9        9           
  Lines        1344     1344           
=======================================
  Hits         1335     1335           
  Misses          9        9           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[dvdplm]

LGTM. My only complaint is about naming.

[andrewwhitehead]

Looks like another 10-15% improvement with the crypto-bigint master. I see a few places (in the Lucas test) that might benefit from a Monty::double method.