multisig checks and optimization

[robert-zaremba]

Description

LegacyAminoMultisig improvements

• Added public key uniqueness check -- though not sure if we want it, so I can remove it.
• Introduced the CompactBitArray.Equal method to improve code flow in x/auth and not use structure fields. (PS: ideally I wanted to make the CompactBitArray attributes private, but the structure is generated with protobuf)
• Optimized the CompactBitArray.NumTrueBitsBefore which is used on every signature validation.
• Added more tests

---

Before we can merge this PR, please make sure that all the following items have been
checked off. If any of the checklist items are not applicable, please leave them but
write a little note why.

• Targeted PR against correct branch (see CONTRIBUTING.md)
• Linked to Github issue with discussion and accepted design OR link to spec that describes this work.
• Code follows the module structure standards.
• Wrote unit and integration tests
• Updated relevant documentation (docs/) or specification (x/<module>/spec/)
• Added relevant godoc comments.
• Added a relevant changelog entry to the Unreleased section in CHANGELOG.md
• Re-reviewed Files changed in the Github PR explorer
• Review Codecov Report in the comment section below once CI passes

[codecov]

Codecov Report

Merging #8600 (2272eaa) into master (8db9bbb) will increase coverage by 0.00%.
The diff coverage is 90.47%.

@@           Coverage Diff           @@
##           master    #8600   +/-   ##
=======================================
  Coverage   61.47%   61.47%           
=======================================
  Files         659      659           
  Lines       37916    37928   +12     
=======================================
+ Hits        23308    23318   +10     
- Misses      12168    12169    +1     
- Partials     2440     2441    +1     

Impacted Files	Coverage Δ
crypto/types/compact_bit_array.go	75.00% <89.47%> (+1.00%)	⬆️
crypto/keys/multisig/multisig.go	71.01% <100.00%> (ø)

[alessio]

It's not worth to improve something that is

1. Broken by design.
2. Fated to be replaced in the near future.

Please close this PR.

[robert-zaremba]

OK, I removed the uniqueness check and updated the comments in the functions to make it clear the not unique public keys are supported.

[alessio]

I'm subscribing @ValarDragon. IIRC he's the original author of these lines.

[robert-zaremba]

I created a benchmark for the optimized function:

BenchmarkNumTrueBitsBefore/old-4         	 4124839	       285 ns/op
BenchmarkNumTrueBitsBefore/new-4         	 7612244	       154 ns/op

So 45% improvement.

[alessio]

I created a benchmark for the optimized function:

BenchmarkNumTrueBitsBefore/old-4         	 4124839	       285 ns/op
BenchmarkNumTrueBitsBefore/new-4         	 7612244	       154 ns/op

So 45% improvement.

Great! I'd love to replicate the benchmarks, do you have code to share please?

[alessio]

Conflicts must be resolved

Lifiting block. Thought I'd like the original author of multisig to review before approving.

[robert-zaremba]

Conflicts must be resolved

I know, there were no conflicts before, I firstly want to have clearance on merging before going back to updating conflicts.

[robert-zaremba]

@alessio , @ValarDragon - could you approve the PR ?

[robert-zaremba]

@alessio - you were upgrading linter recently. It seams that it breaks existing code. Can you confirm?

[ValarDragon]

LGTM

[alessio]

@alessio - you were upgrading linter recently. It seams that it breaks existing code. Can you confirm?

Yes, this can be merged nonetheless. Linter warnings seem minor.

[alessio]

Requested a small change, plus an explanation.

[alessio]

Approving this on the assumption that this PR's author performed all relevant manual testing.