Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Register Assignment Technique in Computer Emulation

IP.com Disclosure Number: IPCOM000111062D
Original Publication Date: 1994-Feb-01
Included in the Prior Art Database: 2005-Mar-26
Document File: 2 page(s) / 44K

Publishing Venue

IBM

Related People

Gohda, O: AUTHOR [+2]

Abstract

Disclosed is a method of assigning registers in simulating a computer by another computer. By mapping one simulated computer's resisters to more than one simulating computer's registers, the simulation can be done more efficiently.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 64% of the total text.

Register Assignment Technique in Computer Emulation

      Disclosed is a method of assigning registers in simulating a
computer by another computer.  By mapping one simulated computer's
resisters to more than one simulating computer's registers, the
simulation can be done more efficiently.

      One way of simulating a computer by another computer is a
method called simulation-time translation.  In the simulation-time
translation method, a block of instructions in the program composed
of the simulated computer's instructions, when that block is to be
simulated, are first translated into the equivalent block composed of
simulating computer's instructions at simulation-time, and then the
translated program is execute on the simulating computer.

      In the above translation process, the simulated computer's
registers are generally mapped to the simulating computer's registers
in one-to-one manner, that is, a simulated computer's register, say,
RA1 is mapped to a simulating computer's register RB1, and so forth.
This one- to-one mapping, however, does not fully utilize the ability
of the simulating computer.  For example, if the simulating computer
has more registers and function units than the simulated computer
has, the one-to-one mapping might not take these advantages.

      By relaxing this one-to-one mapping, the translated instruction
can be executed more efficiently compared to when the registers are
one-to-one mapped.  For example, if the simul...