This paper presents a new assignment and scheduling technique for Hardware/Software partitioning in real-time embedded systems. In contrast to previous approaches, resource constraints are taken into account as soon as possible in order to reduce both the blocking time and the schedulability loss due to resource contention. The system is composed of an arbitrary number of tasks with hard real-time deadlines and different periods. Our method provides the assignment and the scheduling to hardware/software components for each sub-task of each task. Results show that our approach works for cases which cannot be dealt with other methods.