We propose a two-stage gait pattern generation scheme for the full-scale humanoid robots, that considers the dynamics of the system throughout the process. The fist stage is responsible for generating the preliminary motion reference, such as step position, timing and trajectory of Center of Mass (CoM), while the second stage serves as dynamics filter which employs a multi-body model and based on Zero Moment Point (ZMP) reference trajectory generates CoM reference trajectory which defines a locomotion stable when executed on the full-scale multi-degree-of-freedom humanoid robot. The approach thanks to introducing a dynamic model at the stage of feet placement planning provides the ZMP reference, which is ensured to be feasible for the robot. Thanks to the fact it enables instantaneous regeneration of motion. The paper contains description of two approaches used in the first and second stage, as well as experimental results proving the effectiveness of the method. The fast calculation time and the use of the systems dynamic state as initial conditions for pattern generation makes it a good candidate for the real-time gait pattern generator.