Program resource usage amount automatic analysis method based on numerical abstraction

Abstract

The invention discloses a program resource usage amount automatic analysis method based on numerical abstraction. The method comprises the steps of S1, configuration of a resource operation API; S2, code automatic instrumentation; S3, resource reference variable pointing analysis, namely, obtaining a base address set pointed by each resource reference variable in the program and a corresponding resource size through pointing analysis, slicing the instrumented code by utilizing a numerical variable for modeling the resource usage amount and a dependency relationship of the numerical variable, and automatically rejecting codes irrelevant to resource operation to obtain a numerical program; S4, numerical program analysis based on numerical abstraction; S5, cycle index upper bound analysis; and S6, mapping from the numerical program analysis result to the resource usage amount, namely, mapping the numerical program analysis result obtained by analysis to the value range of the resource usage amount. The method has the advantages of simple principle, wide application range, high analysis efficiency, good analysis reliability and the like.

Description

technical field [0001] The invention mainly relates to the technical field of software credibility assurance in computer systems, in particular to an automatic analysis method for program resource usage based on numerical abstraction. Background technique [0002] In safety-critical embedded systems represented by aerospace, etc., the resources that programs can consume (such as heap memory, stack memory, handles, etc.) are often limited by factors such as the system environment and hardware resources. When the resources requested by the program exceed the free resources that the computer can allocate, the running efficiency of the program will be reduced, and it may even cause the exhaustion of computer resources and cause a system failure. Therefore, analyzing the program before running the program, conservatively estimating the maximum amount (peak value) of a certain resource that may be consumed during the running of the program, and assisting in the design of the margi...

AI Technical Summary

Problems solved by technology

The results obtained by the analysis are not reliable upper bounds on resource usage

[0009] 2. Only specific resource types are supported: Currently, the upper bound analysis method of program resources mainly considers heap memory, but many other types of resources are consumed in actual programs, such as file handles, network sockets, etc.

[0010] 3. Low analysis efficiency: Existing analysis techniques use path-sensitive symbolic execution methods, which need to use constraint solving techniques to calculate resource usage for each symbolic path

However, symbolic execution has a path explosion problem and the cost of constraint solving is high, so the analysis efficiency based on symbolic execution is low

Images