Let's now look at one basic result concerning linear programming.
If we optimize a linear function cx over a polyhedron then one of the following situations occur. Either the problem is infeasible, or the problem is unbounded, that means that there is a direction of P along which the function cx strictly increases, or at least one optimal solution of the problem is a vertex of P.