Constrained multi-objective optimization problems (CMOPs) encompass multiple conflicting objectives and a multitude of intricate constraints. Solving CMOPs requires balancing between objectives and constraints, which can be a demanding task due to the intricate and irregular geometry of the feasible region. To confront the previously mentioned challenge, this study has introduced a constrained multi-objective differential evolutionary algorithm based on evolutionary multi-task optimization (CMODE-EMO). The proposed CMODE-EMO employs a main task and two auxiliary tasks. The main task is centered on addressing the original CMOPs, while the first auxiliary task considers objectives without constraints, and the second auxiliary task emphasizes the feasibility of solutions. Moreover, a self-adaptive individual transfer strategy is proposed to facilitate knowledge exchange among various tasks. The proposed algorithm is assessed across 15 CMOPs, and its effectiveness is measured through a comparison with seven established and advanced CMOEAs. The experimental findings clearly illustrate that CMODE-EMO outperforms all other competitors in terms of overall performance, establishing that the proposed multi-task constraint multi-objective optimization approach is highly effective and competitive.