Reuse root context in a sync pool to reduce GC: zero-alloc routing

[pkieltyka]

This PR removes allocations during routing and speeds up overall performance.

Note: there is one breaking change, which is how URL params are referenced. Previously a handler would reference chi.URLParams(ctx)["id"] and now, its: chi.URLParam(ctx, "id")

New Benchmarks:

BenchmarkChi_Param          10000000           181 ns/op           0 B/op          0 allocs/op
BenchmarkChi_Param5          3000000           570 ns/op           0 B/op          0 allocs/op
BenchmarkChi_Param20         1000000          2057 ns/op           0 B/op          0 allocs/op
BenchmarkChi_ParamWrite      5000000           245 ns/op           0 B/op          0 allocs/op
BenchmarkChi_GithubStatic    5000000           250 ns/op           0 B/op          0 allocs/op
BenchmarkChi_GithubParam     2000000           589 ns/op           0 B/op          0 allocs/op
BenchmarkChi_GithubAll         10000        102664 ns/op           0 B/op          0 allocs/op
BenchmarkChi_GPlusStatic    10000000           161 ns/op           0 B/op          0 allocs/op
BenchmarkChi_GPlusParam      5000000           291 ns/op           0 B/op          0 allocs/op
BenchmarkChi_GPlus2Params    5000000           393 ns/op           0 B/op          0 allocs/op
BenchmarkChi_GPlusAll         300000          4335 ns/op           0 B/op          0 allocs/op
BenchmarkChi_ParseStatic    10000000           162 ns/op           0 B/op          0 allocs/op
BenchmarkChi_ParseParam     10000000           227 ns/op           0 B/op          0 allocs/op
BenchmarkChi_Parse2Params    5000000           327 ns/op           0 B/op          0 allocs/op
BenchmarkChi_ParseAll         200000          7368 ns/op           0 B/op          0 allocs/op
BenchmarkChi_StaticAll         30000         57990 ns/op           0 B/op          0 allocs/op

Old Benchmarks:

BenchmarkChi_Param           2000000           821 ns/op         408 B/op          6 allocs/op
BenchmarkChi_Param5          1000000          1478 ns/op         408 B/op          6 allocs/op
BenchmarkChi_Param20          200000          6595 ns/op        2259 B/op          8 allocs/op
BenchmarkChi_ParamWrite      2000000           954 ns/op         416 B/op          7 allocs/op
BenchmarkChi_GithubStatic    3000000           585 ns/op         120 B/op          5 allocs/op
BenchmarkChi_GithubParam     1000000          1286 ns/op         408 B/op          6 allocs/op
BenchmarkChi_GithubAll         10000        237874 ns/op       72456 B/op       1182 allocs/op
BenchmarkChi_GPlusStatic     3000000           495 ns/op         120 B/op          5 allocs/op
BenchmarkChi_GPlusParam      2000000           945 ns/op         408 B/op          6 allocs/op
BenchmarkChi_GPlus2Params    1000000          1100 ns/op         408 B/op          6 allocs/op
BenchmarkChi_GPlusAll         100000         12993 ns/op        4728 B/op         76 allocs/op
BenchmarkChi_ParseStatic     3000000           503 ns/op         120 B/op          5 allocs/op
BenchmarkChi_ParseParam      2000000           902 ns/op         408 B/op          6 allocs/op
BenchmarkChi_Parse2Params    1000000          1029 ns/op         408 B/op          6 allocs/op
BenchmarkChi_ParseAll         100000         21745 ns/op        7728 B/op        146 allocs/op
BenchmarkChi_StaticAll         10000        110299 ns/op       18840 B/op        785 allocs/op

[pkieltyka]

btw.. for fun I ran the benchmarks on Go-tip (1.7)

BenchmarkChi_Param                  10000000           143 ns/op           0 B/op          0 allocs/op
BenchmarkChi_Param5                  5000000           368 ns/op           0 B/op          0 allocs/op
BenchmarkChi_Param20                 1000000          1336 ns/op           0 B/op          0 allocs/op
BenchmarkChi_ParamWrite             10000000           196 ns/op           0 B/op          0 allocs/op
BenchmarkChi_GithubStatic           10000000           220 ns/op           0 B/op          0 allocs/op
BenchmarkChi_GithubParam             3000000           445 ns/op           0 B/op          0 allocs/op
BenchmarkChi_GithubAll                 20000         75896 ns/op           0 B/op          0 allocs/op
BenchmarkChi_GPlusStatic            10000000           140 ns/op           0 B/op          0 allocs/op
BenchmarkChi_GPlusParam             10000000           200 ns/op           0 B/op          0 allocs/op
BenchmarkChi_GPlus2Params            5000000           287 ns/op           0 B/op          0 allocs/op
BenchmarkChi_GPlusAll                 500000          3036 ns/op           0 B/op          0 allocs/op
BenchmarkChi_ParseStatic            10000000           145 ns/op           0 B/op          0 allocs/op
BenchmarkChi_ParseParam             10000000           180 ns/op           0 B/op          0 allocs/op
BenchmarkChi_Parse2Params           10000000           234 ns/op           0 B/op          0 allocs/op
BenchmarkChi_ParseAll                 300000          5351 ns/op           0 B/op          0 allocs/op
BenchmarkChi_StaticAll                 30000         47330 ns/op           0 B/op          0 allocs/op