Crosspoint buffered crossbar switches have gained much attention due to simple distributed scheduling algorithms. However, almost all the algorithms proposed for buffered crossbar switches either have unsatisfactory scheduling performance under non-uniform traffic or poor service fairness between input traffic flows. In order to overcome the disadvantages, in this paper we propose a novel efficient scheduling algorithm named MCQF_RR, which takes advantage of the combined information about queue length and service waiting time of queues to perform scheduling, in order to ensure good service fairness. Simulation results show that the proposed MCQF_RR demonstrates good delay performance comparable to existing efficient algorithms under various admissible traffic patterns, while achieve better service fairness under extreme nonuniform traffic.