mio-multithread-unix

This is the same as the mio-singlethread benchmark, except it leverages SO_REUSEPORT and SO_REUSEADDR to have multiple event loops all listening on the same socket. This is currently a Unix-specific option and isn't available on Windows.

The performance should be greater than mio-singlethread on multicore machines, and otherwise the code is intended to be the same.

See the mio-singlethread README for how to run