Benchmark get canonical validator set

[aaronbuchwald]

Why this should be merged

This PR adds a benchmark for GetCanonicalValidatorSet used in warp signature verification.

How this works

See the code

How this was tested

Running locally to confirm the benchmark passes and see the initial values:

# github.com/ava-labs/avalanchego/vms/platformvm/warp.test
ld: warning: could not create compact unwind for _blst_sha256_block_data_order: does not use RBP or RSP based frame
goos: darwin
goarch: amd64
pkg: github.com/ava-labs/avalanchego/vms/platformvm/warp
cpu: Intel(R) Core(TM) i7-9750H CPU @ 2.60GHz
BenchmarkGetCanonicalValidatorSet/0-12  	37213069	        29.15 ns/op	       0 B/op	       0 allocs/op
BenchmarkGetCanonicalValidatorSet/1-12  	 2807650	       441.2 ns/op	     192 B/op	       5 allocs/op
BenchmarkGetCanonicalValidatorSet/10-12 	  275293	      5106 ns/op	    2340 B/op	      42 allocs/op
BenchmarkGetCanonicalValidatorSet/100-12         	   24452	     54903 ns/op	   23477 B/op	     403 allocs/op
BenchmarkGetCanonicalValidatorSet/1000-12        	    2030	    510480 ns/op	  249560 B/op	    4003 allocs/op
BenchmarkGetCanonicalValidatorSet/10000-12       	     184	   6788629 ns/op	 2380698 B/op	   40003 allocs/op
PASS
ok  	github.com/ava-labs/avalanchego/vms/platformvm/warp	12.603s

[aaronbuchwald]

@joshua-kim this is the PR adding the benchmark for GetCanonicalValidatorSet. I separated the function out locally and saw that the lion's share of the time was spent handling the sort operation at the end.

Your idea of adding the items to a sorted map sounds like a good idea to optimize this.