Use RandomSource in a few more places

[fingolfin]

This PR makes some progress on issue #1098.

[codecov]

Codecov Report

Merging #1599 into master will increase coverage by 0.02%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master    #1599      +/-   ##
==========================================
+ Coverage   64.45%   64.47%   +0.02%     
==========================================
  Files        1003     1002       -1     
  Lines      326890   326805      -85     
  Branches    13216    13239      +23     
==========================================
+ Hits       210696   210708      +12     
+ Misses     113331   113234      -97     
  Partials     2863     2863

Impacted Files	Coverage Δ
lib/modfree.gi	78.76% <100%> (ø)	⬆️
lib/random.gi	76.41% <100%> (+0.45%)	⬆️
lib/smgrpfre.gi	74.77% <100%> (+4.86%)	⬆️
hpcgap/lib/zmodnz.gi	93.84% <100%> (ø)	⬆️
hpcgap/lib/random.g	90.47% <100%> (+27.31%)	⬆️
lib/matrix.gi	76.88% <100%> (+0.32%)	⬆️
lib/zmodnz.gi	95.41% <100%> (ø)	⬆️
lib/fieldfin.gi	63.7% <100%> (ø)	⬆️
lib/queue.g	66.4% <0%> (-3.2%)	⬇️
src/vars.h	93.93% <0%> (-3.04%)	⬇️
... and 20 more

[fingolfin]

I rebased this, fixed the test failures (I hope, we'll see when Travis completed) and added the previously missing documentation update for the RandomMat functions.

Hopefully this can be merged soon -- @ChrisJefferson if you have any thoughts on this PR, please ... :-)

[fingolfin]

Turns out initialization of the old-style GlobalRandomSource in HPC-GAP was subtly broken. So I added another commit to fix that, and while at it, also merged hpcgap/lib/random.gi into lib/random.gi. Let's see if tests pass now.