Tailored for the best effort service, the present paper deals with dynamic allocation of subcarriers, rate and power resources based on channel state information (CSI) for multiuser MIMO-OFDM systems with trellis code modulation (TCM). Users rely on adaptive modulation and power modes that they select in accordance with the finite-rate feedback (FRF) they receive from the base station (BS). The BS uses CSI to maximize a generic concave utility of average rates while adhering to power constraints imposed on every user. We utilize a stochastic approximation primal-dual algorithm for on-line scheduling and resource allocation. The resultant optimum resource allocation depends on the current channel realization and optimally calculates dual prices. The simulation results show that the system throughput of proposed algorithm is not only good but also can effectively reduce the feedback overhead and the adoption of TCM can improve the system throughput greatly comparing to uncoded system.