The control barrier function (CBF) is used as an effective tool to design the safety controls for constrained robotic systems. When both the control input and state constraints are involved, it is important to guarantee the CBF-based quadratic program (QP) is feasible. To address the constraint satisfaction and CBF-QP feasibility issues for robotic systems with saturation constraints both in control input and velocity state, in this article, we propose an augment CBF (ACBF), which is constructed by introducing a time-varying function to a typical ZCBF, and the time-varying function is treated as a ZCBF by introducing the auxiliary dynamics. To realize the safe tracking purpose for the robotic systems, the ACBF-based QP tracking control is designed by applying the computed torque control (CTC) as a nominal tracking control and the designed ACBF condition as the constraint of the QP problem. The performance of proposed ACBF-QP control method is illustrated by a two-link manipulator system.