MSM CUDA implementation

[Protryon]

This PR adds a CUDA MSM implementation. Note that CI does NOT test this feature, since it relies on a GPU being present.

[codecov]

Codecov Report

Merging #203 (5bb7343) into master (c2d080c) will decrease coverage by 0.02%.
The diff coverage is n/a.

❗ Current head 5bb7343 differs from pull request most recent head cfa660a. Consider uploading reports for the commit cfa660a to get more accurate results

@@            Coverage Diff             @@
##           master     #203      +/-   ##
==========================================
- Coverage   93.16%   93.14%   -0.03%     
==========================================
  Files          60       56       -4     
  Lines        2679     2670       -9     
==========================================
- Hits         2496     2487       -9     
  Misses        183      183              

Impacted Files	Coverage Δ
curves/src/templates/bls12/bls12.rs	95.23% <0.00%> (-2.39%)	⬇️
fields/src/fp_256.rs	91.57% <0.00%> (-1.06%)	⬇️
fields/src/fp_768.rs	91.56% <0.00%> (-0.51%)	⬇️
utilities/src/bytes.rs	91.66% <0.00%> (-0.34%)	⬇️
fields/src/macros.rs	98.03% <0.00%> (-0.08%)	⬇️
wasm/src/lib.rs
r1cs/src/lib.rs
.integration/src/lib.rs
r1cs/src/linear_combination.rs
fields/src/fp6_2over3.rs	97.59% <0.00%> (+0.02%)	⬆️
... and 2 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update c2d080c...cfa660a. Read the comment docs.