A static detection method for state-related defects in a large-scale system

Abstract

The invention relates to a static detection method for state-related defects in a large-scale system. The static detection method comprises the following steps: firstly, abstracting the state-relateddefects in the large-scale system into a form represented by a finite state machine; secondly, generating an inter-process control flow execution tree (ICFET) based on a symbol execution technology, and proposing a new interval-based path encoding/decoding algorithm to represent path constraints; based on the ICFET, alias analysis and data flow analysis are then performed in a context-sensitive and path-sensitive manner, tracking the flow of each object of each specified type to identify a sequence of events that may occur on the object. And finally, generating a program diagram according to the ICFET and alias analysis and data flow analysis. Constraint-based graph reachability calculation is performed, and if a reachable edge can reach an undefined or erroneous state in a finite state machine, there is an event sequence of undefined or erroneous states that can drive an object or group of objects to a specification, generating an erroneous report.

Description

technical field [0001] The invention relates to a static detection method for state-related defects in large-scale systems, and belongs to the application fields of software engineering, computer systems, and high-performance computing intersecting technologies. Background technique [0002] Large-scale software systems, including operating systems, web browsers, databases, database processing engines, etc., constitute the backbone of modern computing. Since these systems are widely used in various fields, ensuring their reliability is of paramount importance. Despite continuous efforts by industry and academia to make these systems more reliable, errors still occur frequently in large-scale software systems. [0003] A common type of error in large-scale software systems is state-dependent: after a finite sequence of events has occurred to an object of interest, an error occurs that drives the object into a wrong state. Examples include acquiring locks that are not releas...

AI Technical Summary

Problems solved by technology

[0006] Technical problem: The present invention provides a static detection method for state-related defects in large-scale systems, the purpose of which is to solve the problem that existing technologies are difficult to detect state-related defects in large-scale systems, and can perform limited execution on very large code bases. Accurate and scalable static checking of state properties, efficiently and accurately detecting a variety of errors in large-scale systems, including source-sink problems (such as resource leaks), exception handling errors (such as missing error handling code), or type status related errors (such as inappropriate use of file handlers), etc.

Images