Method for refining static defect detection on basis of state partitions

Abstract

The invention relates to a method for refining static defect detection on the basis of state partitions. The method includes S1, placing sets of to-be-detected defect mode property state machines in to-be-detected function control flow inlets; S2, executing programs from front locations to target locations via paths; S3, transmitting states of the property state machines along the paths of the programs, changing the states of the property state machines and enabling the property state machines to be in preset states at the target locations; S4, acquiring current control flow nodes and judging whether the current control flow nodes are final control flow nodes or not; S5, updating property state conditions and information of the sets of the property state partitions according to information of the current nodes if the current nodes are not the final control flow nodes. The method for refining static defect detection on the basis of a technology of the state partitions has the advantages that defect modes, the defect mode state machines, path conditions, the state conditions and the state partitions are defined, the method for refining defect detection is provided on the basis of the technology of the state partitions, and accordingly the defect detection precision can be improved.

Description

technical field [0001] The invention relates to a detection and refinement technology in static defect detection of computer software, in particular to a state partition-based static defect detection and refinement method. Background technique [0002] Flaws in code are a significant cause of software failures and security breaches. Defect-based software testing technology can be divided into dynamic detection technology and static detection technology. [0003] Among them, static detection technology does not run the program under test, and mainly uses various static analysis methods to find defects in the program. From the perspective of computability theory, static analysis is an undecidable problem. The undecidability of static analysis is actually This means that any automated static analysis system cannot be both reliable and complete for non-trivial properties of a program, such as whether there are runtime errors. Reliable static analysis means that if the analysis...

AI Technical Summary

Problems solved by technology

[0004] The path-sensitive detection method based on data flow analysis considers the combination relationship between program path branches, and can record different path information on the control flow graph, thereby effectively reducing false positives during static analysis. The accurate path-sensitive analysis method will record the program. All path information will lead to path explosion when there are many control flow branches or loops, so that it cannot be analyzed. Therefore, practical path-sensitive analysis methods

Images