Fast permutation vector: code needed

General Usage Performance
10

Are you measuring with something like:

julia> const v = rand(10_000);

julia> const p = zeros(Int, 10_000);

julia> using BenchmarkTools

julia> @btime sortperm!($p, $v);
  666.553 μs (1 allocation: 16 bytes)

(use bigger vector if 10_000 small)
?