Morphing quadrotors with four external actuators can adapt to different restricted scenarios by changing their geometric structure. However, previous works mainly focus on the improvements in structures and controllers, and existing planning algorithms don't consider the morphological modifications, which leads to safety and dynamic feasibility issues. In this paper, we propose a unified planning and control framework for morphing quadrotors to deform autonomously and efficiently. The framework consists of a milliseconds-level spatial-temporal trajectory optimizer that takes into account the morphological modifications of quadrotors. The optimizer can generate full-body safety trajectories including position and attitude. Additionally, it incorporates a nonlinear attitude controller that accounts for aerodynamic drag and dynamically adjusts dynamic parameters such as the inertia tensor and Center of Gravity. The controller can also online compute the thrust coefficient during morphing. Benchmark experiments compared with existing methods validate the robustness of the proposed controller. Extensive simulations and real-world experiments are performed to demonstrate the effectiveness of the proposed framework.