Inverse Reinforcement Learning Meets Large Language Model Post-Training: Basics, Advances, and Opportunities