Cycle Count on ARM Cortex A9 or A15 Series or related Board

We have algorithms (programs written in C Language) that take variable length input and generate fixed size output (to be precise these are hash functions / algorithms). we need to run these programs on ARM Cortex A9 or A15 (any one) or such other Board. The purpose is to evaluate the clock cycles taken to run the program for set of different inputs. Important point : a) For evaluating the performance rather than using execution speed in units of time, we want 'Cycles per Byte’. ‘Cycles per Byte’ is architecture oriented and will not change with the frequency of the device. b) We may need cycle count for the whole program as well as for a few specific functions (subroutines) written in the program. Total number and type of inputs for which the same is to be calculated can be mutually decided. One of the program is attached herewith. The other programs can be provided after the contract finalizes.