Unbounded test for linear continuous

[matbesancon]

See jump-dev/Clp.jl#60
The unbounded case seems untested, at least for linear problems

[blegat]

I agree, we should also add a constant in the objective and check that it is ignored in ObjectiveValue for the infeasibility ray (see #433)

[matbesancon]

A non-tested example was added in jump-dev/Clp.jl#60