refactor: use `stdlib_base_fmod` instead of built-in in `math/base/special/gcd`

[gunjjoshi]

Description

What is the purpose of this pull request?

This pull request:

• replaces the use of built-in fmod by our stdlib_base_fmod in math/base/special/gcd.

Related Issues

Does this pull request have any related issues?

None.

Questions

Any questions for reviewers of this pull request?

No.

Other

Any other information relevant to this pull request? This may include screenshots, references, and/or implementation notes.

I've changed the js implementation as well, to use our fmod instead of its % operator, so that both the implementations are close enough.

Checklist

Please ensure the following tasks are completed before submitting this pull request.

• [x] Read, understood, and followed the contributing guidelines.

---

@stdlib-js/reviewers

[gunjjoshi]

As discussed, I generated benchmark results before and after making the changes proposed in this PR.

Before:

benchmark.js:

TAP version 13
# @stdlib/math/base/special/gcd
  ---
  iterations: 10000000
  elapsed: 0.202160416
  rate: 49465667.89811117
  ...
ok 1 benchmark finished
# @stdlib/math/base/special/gcd
  ---
  iterations: 10000000
  elapsed: 0.200456208
  rate: 49886207.565095715
  ...
ok 2 benchmark finished
# @stdlib/math/base/special/gcd
  ---
  iterations: 10000000
  elapsed: 0.200792958
  rate: 49802543.37405598
  ...
ok 3 benchmark finished
#
1..3
# total 3
# pass  3
#
# ok

benchmark.native.js:

TAP version 13
# @stdlib/math/base/special/gcd::native
  ---
  iterations: 10000000
  elapsed: 0.446446833
  rate: 22399083.74487226
  ...
ok 1 benchmark finished
# @stdlib/math/base/special/gcd::native
  ---
  iterations: 10000000
  elapsed: 0.442721583
  rate: 22587559.278762337
  ...
ok 2 benchmark finished
# @stdlib/math/base/special/gcd::native
  ---
  iterations: 10000000
  elapsed: 0.452107708
  rate: 22118623.113587793
  ...
ok 3 benchmark finished
#
1..3
# total 3
# pass  3
#
# ok

After:

benchmark.js:

TAP version 13
# @stdlib/math/base/special/gcd
  ---
  iterations: 10000000
  elapsed: 0.205398375
  rate: 48685876.89654312
  ...
ok 1 benchmark finished
# @stdlib/math/base/special/gcd
  ---
  iterations: 10000000
  elapsed: 0.202361708
  rate: 49416463.711602986
  ...
ok 2 benchmark finished
# @stdlib/math/base/special/gcd
  ---
  iterations: 10000000
  elapsed: 0.207460667
  rate: 48201908.075423285
  ...
ok 3 benchmark finished
#
1..3
# total 3
# pass  3
#
# ok

benchmark.native.js:

TAP version 13
# @stdlib/math/base/special/gcd::native
  ---
  iterations: 10000000
  elapsed: 0.435659958
  rate: 22953681.687679913
  ...
ok 1 benchmark finished
# @stdlib/math/base/special/gcd::native
  ---
  iterations: 10000000
  elapsed: 0.433399375
  rate: 23073406.601013213
  ...
ok 2 benchmark finished
# @stdlib/math/base/special/gcd::native
  ---
  iterations: 10000000
  elapsed: 0.438695959
  rate: 22794830.43973058
  ...
ok 3 benchmark finished
#
1..3
# total 3
# pass  3
#
# ok

[gunjjoshi]

Looks like while the addon performance has improved, the performance of javascript became slower. Only changing the C implementation could be a possible way.

[kgryte]

@gunjjoshi Yes, I think only using fmod in the C implementation is the path forward here.