Division improvements

[chfast]

This makes 256-bit division ~40% faster and 512-bit division ~10% slower.

Benchmark                                   Time             CPU      Time Old      Time New       CPU Old       CPU New
------------------------------------------------------------------------------------------------------------------------
udiv<uint256, uint32_t, udivrem>         -0.4528         -0.4528         56945         31162         56945         31162
udiv<uint256, uint64_t, udivrem>         -0.3995         -0.3995         55146         33118         55145         33117
udiv<uint256, uint128, udivrem>          -0.4197         -0.4198         58186         33763         58186         33762
udiv<uint256, uint256, udivrem>          -0.0868         -0.0869         37118         33894         37117         33893
udiv<uint512, uint32_t, udivrem>         +0.0660         +0.0660         56905         60663         56905         60662
udiv<uint512, uint64_t, udivrem>         +0.0833         +0.0833         55982         60648         55982         60647
udiv<uint512, uint256, udivrem>          -0.0463         -0.0463         97535         93020         97533         93018

[codecov-io]

Codecov Report

Merging #99 into master will increase coverage by 2.61%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master      #99      +/-   ##
==========================================
+ Coverage   71.08%   73.69%   +2.61%     
==========================================
  Files           8        8              
  Lines        1349     1346       -3     
  Branches      162      158       -4     
==========================================
+ Hits          959      992      +33     
+ Misses        342      306      -36     
  Partials       48       48