Cloud radio access network (C-RAN) and massive multiple-input-multiple-output (MIMO) are two key enabling technologies to meet the diverse and stringent requirements of the 5G use cases. In a C-RAN system with massive MIMO, fronthaul is often the bottleneck due to its finite capacity and transmit precoding is moved to the remote radio head to reduce the capacity requirements on fronthaul. For such a system, we optimize the power allocated to the users to maximize first the weighted sum rate and then the energy efficiency (EE) while explicitly incorporating the capacity constraints on fronthaul. We consider two different fronthaul constraints, which model capacity constraints on different parts of the fronthaul network. We develop successive convex approximation algorithms that achieve a stationary point of these non-convex problems. To this end, we first present novel, locally tight bounds for the user rate expression. They are used to obtain convex approximations of the original non-convex problems, which are then solved by solving their dual problems. In EE maximization, we also employ the Dinkelbach algorithm to handle the fractional form of the objective function. Numerical results show that the proposed algorithms significantly improve the network performance compared to a case with no power control and achieves a better performance than an existing algorithm.