This thread, especially the top post, already contains the Benchmark and the usual scaffolding so you can easily test the speed (and correctness) of your implementation. Maybe consider using that.

Also, I highly doubt that an implementation of Quicksort will beat the built-in sort keyword at least for the short lists that we are talking about here. Even though Perl sort uses a Mergesort, it will be hard for Perl code to beat the C code simply due to the amount of time spent in the Perl runloop.