Random code generation using genetic algorithms

Abstract

Techniques are disclosed for automatically generating test instructions for use in testing a microprocessor design. A configuration file includes a plurality of knobs which specify a probability distribution of a plurality of microprocessor instructions. A random code generator takes the configuration file as an input and generates test instructions which are distributed according to the probability distribution specified by the knobs. The test instructions are executed on the microprocessor design. The microprocessor behaviors that are exercised by the test instructions are measured and a fitness value is assigned to the configuration file using a fitness function. The configuration file and its fitness value are added to a pool of configuration files. A configuration file synthesizer uses a genetic algorithm to synthesize a new configuration file from the pool of existing configuration files. This process may be repeated to generate configuration files which increasingly exercise microprocessor behaviors which are of interest.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS [0001] This application is a divisional application of commonly-owned U.S. patent application Ser. No. 10 / 195,993, filed on Jul. 16, 2002, entitled “Random Code Generation Using Genetic Algorithms”.BACKGROUND [0002] 1. Field of the Invention [0003] The present invention relates to microprocessor design and, more particularly, to the automatic generation of test instructions for microprocessor designs. [0004] 2. Related Art [0005] Various electronic design automation (EDA) software tools exist for designing microprocessors and other circuitry. Such tools allow circuit designers to create and modify virtual models of the circuit being designed. A circuit designer may, for example, specify a circuit design using a textual description written in a hardware description language (HDL), such as Verilog or VHDL, or by using a graphical user interface to manipulate a graphical representation of the circuit design. [0006] Software tools are also frequen...

AI Technical Summary

Benefits of technology

"The present invention provides a method for automatically generating test instructions for a microprocessor design. The method involves identifying a plurality of knobs that specify the probability distribution of a plurality of circuit input vectors. These knobs are then used to generate test instructions that exercise the microprocessor design. The test instructions are measured and a fitness value is assigned to each knob. A pool of configuration files is created based on the fitness values. A genetic algorithm is used to synthesize new configuration files by combining the pool of configuration files. The technical effect of this invention is to automate the process of creating test instructions for a microprocessor design, making it faster and more efficient."

Problems solved by technology

The execution of particular test instructions, or the results produced by executing such instructions, may cause one or more events to occur.

It is difficult, however, for the circuit designer 116 to design the configuration file 108 to test (or “exercise”) the entire range of events or particular events of interest because, for example, the relationship between the execution of particular instruction sequences and the occurrence of particular events may be complex and difficult to predict.

Although the circuit designer 116 may improve the extent to which the configuration file 108 exercises particular events by inspecting the event counters 106 and / or the simulation results 118 upon completion of the simulation performed by the simulator 114 and modifying the configuration file 108 manually in response, such a process is tedious, time-consuming, and not guaranteed to improve the extent to which the configuration file 108 exercises events of interest.

Images