Migrate mantle tests to use magma simulator

[leonardt]

Should be more robust than regression testing the generator verilog files

[shacklettbp]

My strategy is currently just to simulate small circuits and verify the simulated outputs are consistent with the inputs, does this seem reasonable? I don't look at the verilog at all.

[shacklettbp]

Ah nevermind, I misinterpreted what this issue was saying.