Memory Leak Detection in IoT Program Based on an Abstract Memory Model SeqMM

Yukun Dong,Li Zhang,Shudong Wang,Wenjing Yin,Lin Sun

doi:10.1109/access.2019.2951168

Abstract

Memory Leak Detection in IoT Program Based on an Abstract Memory Model SeqMM

Highlights

As the basic infrastructure of the information society, the IoT has been applied in many fields, such as medical monitoring, intelligent transportation, and environmental monitoring
In previous work, we proposed a sound abstract memory model region-based symbolic three-valued logic(RSTVL) [5], which can describe the morphological information of data structures in memory and the storage state of the memory objects of C programs
We have proposed a sound abstract memory model RSTVL [5] that can describe points-to relationships, hierarchical relationships, linear and logic relationships among memory addressable expressions but does not support the description of the pointer offset, leading to inadequate detection for defects related to sequential storage structures, such as memory leak, out-of-bounds and buffer overflow defects

Summary

INTRODUCTION

As the basic infrastructure of the information society, the IoT has been applied in many fields, such as medical monitoring, intelligent transportation, and environmental monitoring. Based on the study of static analysis methods and memory leak detection for sequential storage structures, this paper makes the following contributions:. In previous work, we proposed a sound abstract memory model region-based symbolic three-valued logic(RSTVL) [5], which can describe the morphological information of data structures in memory and the storage state of the memory objects of C programs. The method recently proposed in [30] dynamically detect memory leaks such that relevant information about each allocated memory block is updated as the program is running. We use static analysis to model the sequential storage structure that may cause memory leaks, and to effectively analyze the various operations involved, especially considering pointer points-to relationships and offsets. Based on the results of the static analysis, we proposed a memory leak detection method and verified it on five open-source projects

DATA FLOW ANALYSIS FOR SEQUENTIAL STORAGE STRUCTURE

INTER-PROCEDURAL ANALYSIS

MEMORY LEAK DETECTION ALGORITHM

THE EXPERIMENT

Findings

MEMORY LEAK DETECTION OF 5 OPEN-SOURCE OBJECTS