FPGATrackSim: Fix sign of FPGATrackSimHit::getEtaModule return value
This function returns the eta module of a hit, which is stored internally as
a signed integer-- because hits on the negative eta side of the detector have
negative eta modules. However the actual function was declared as unsigned...
so it would be returning the unsigned representation of the negative integer.
I think that in the actual codebase we were always implicitly casting this
back to an integer (by e.g. int eta = hit.getEtaModule()), which is why
it didn't get noticed sooner. We noticed it when preparing test vectors from
the ROOT output of FPGATrackSim (which stores these classes in a TTree, and
Python code is used to load the objects and spit out FPGA-friendly formats).
This should be fixed.