Add benchmarks

[fjarri]

Fixes #5

Add some benchmarks to have a rough idea of how much overhead our framework introduces. These are all-encompassing, checking pretty much all of Session's happy paths. This will gives us a general idea when some regressions happen during code refactoring and adding new features. More specific benchmarks can be added later.

On my machine, 25 nodes/5 rounds without echo takes 0.34ms per node, and with echo 5ms per node. The first number will scale linearly as the number of node increases (so the whole test time will scale quadratically), while the second one will scale quadratically (so the whole test time will scale cubically) since the length of each echo round on each node scales linearly with the number of nodes.

[coveralls]

Pull Request Test Coverage Report for Build 11420762300

Details

• 1 of 4 (25.0%) changed or added relevant lines in 2 files are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage decreased (-0.1%) to 75.864%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
manul/src/protocol/round.rs	0	3	0.0%

Totals
Change from base Build 11420136282:	-0.1%
Covered Lines:	1559
Relevant Lines:	2055

---

💛 - Coveralls