When data compression is added to a front-end processor or programmable communications controller, several method can be considered to link the compression module to existing software. The compression software can be added to an existing module or it can be coded as a separate module. For the latter situation, several methods can be used to invoke its operation. If the compression routine is to reside on a minicomputer, a ‘jump and store return address’ instruction or subroutine call are the two most frequently used methods for one routine to invoke another. At the completion of the routine, a return to the invoking instruction plus one location occurs and the instruction following the jump or the call instruction is executed next. If compression is to be employed with microcomputers, hardware and software differences may preclude the use of a jump or call statement. Many microprocessors are designed so that complete functions, such as a data-compression routine, may be burned into chips which are connected to the microprocessor’s I/O or DMA bus. Since these coded chips are then considered part of the microprocessor’s memory, linkage becomes a memory access problem. For these microprocessors that lack a ‘jump and store return address’ or subroutine call instruction, the address of the chip can be stored in one of the processor registers. Then an indirect address through the register will result in the processor fetching the chip address to initiate the compression routine.