Embedded systems typically require both periodic and aperiodic computation and some are driven by user interface whereas some respond to the network. The system scheduler overhead and context switch overhead provide the lower bound on the timer interrupt period. Similarly, the deadlines of the processes, periods of the processes and execution time of the processes provide the upper bound. Hardware-Software Partitioning and decompilation is a key issue in the Codesign of embedded systems. Partitioning in binary level helps in independent usage of software languages for the compilers. In this paper, the critical kernel of the software binary is relocated to the hardware and this is identified using instruction level profiling. The proposed work has the advantage that the throughput gets improved. A generalized decompiler is also designed to generate equivalent HDL for software binary block. Code motion is also applied wherever applicable, such that constant array values in an inner loop, is relocated outside the loop. Unrolling is also done to exploit parallelism and improve speedup. The proposed method is applied to standard benchmark and show significant speedup with lesser hardware resources.