Browse Prior Art Database

Operand Address Calculation Hardware for a Microprocessor

IP.com Disclosure Number: IPCOM000045454D
Original Publication Date: 1983-Mar-01
Included in the Prior Art Database: 2005-Feb-07
Document File: 1 page(s) / 12K

Publishing Venue

IBM

Related People

Buonomo, JP: AUTHOR [+4]

Abstract

The effective emulation of main frame instruction sets in and by a microprocessor (MPU)-based system has been realized. The efficiency of such emulation is often boosted by hardware assists that facilitate manipulations that could not otherwise be effectively performed. One such area in which added hardware may help MPU emulation performance is that of I-cycle operand address calculation.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 82% of the total text.

Page 1 of 1

Operand Address Calculation Hardware for a Microprocessor

The effective emulation of main frame instruction sets in and by a microprocessor (MPU)-based system has been realized. The efficiency of such emulation is often boosted by hardware assists that facilitate manipulations that could not otherwise be effectively performed. One such area in which added hardware may help MPU emulation performance is that of I-cycle operand address calculation.

On-chip hardware is provided that zeros the "B" field of the second or third halfword of an IBM System 370 instruction as it enters the MPU. As a result of this hardware action, the data-in register contains an X'DDD' value after an instruction halfword two or three fetch. The data-in register is then sign extended and sent to the MPU's 24-bit adder unit to form the X'000DDD' component of the operand address. This portion of the operand address calculation takes one machine cycle.

Without the use of this "B" field zeroing capability, the X'000DDD' component of the operand address would have to be formed by some other method or means. One possible method would transfer the X'BDDD' field into the data-in register and then "AND" it with a constant of X'OFFF' in the MPU's ALU. This approach, however, requires constant generation resources and an additional machine cycle over the strategy proposed above. Other solutions that have been considered require considerably more hardware and/or time to run.

In addition to the advantag...