左万娟, 董燕, 黄晨, 王小丽. 航天嵌入式软件静态测试方法研究[J]. 微电子学与计算机, 2022, 39(5): 78-86. DOI: 10.19304/J.ISSN1000-7180.2021.1129
引用本文: 左万娟, 董燕, 黄晨, 王小丽. 航天嵌入式软件静态测试方法研究[J]. 微电子学与计算机, 2022, 39(5): 78-86. DOI: 10.19304/J.ISSN1000-7180.2021.1129
ZUO Wanjuan, DONG Yan, HUANG Chen, WANG Xiaoli. Research on static testing method of aerospace embedded software[J]. Microelectronics & Computer, 2022, 39(5): 78-86. DOI: 10.19304/J.ISSN1000-7180.2021.1129
Citation: ZUO Wanjuan, DONG Yan, HUANG Chen, WANG Xiaoli. Research on static testing method of aerospace embedded software[J]. Microelectronics & Computer, 2022, 39(5): 78-86. DOI: 10.19304/J.ISSN1000-7180.2021.1129

航天嵌入式软件静态测试方法研究

Research on static testing method of aerospace embedded software

  • 摘要: 航天器在轨运行场景高度复杂,部分场景很难在地面实现真实状态下的动态验证.通过对近30年来航天软件在轨、在研以及第三方评测发现缺陷的分析与研究,提出了一套涵盖检查单法、变量分析法、中断访问冲突分析法、代码逻辑分析法、工具静态扫描分析法在内的以人工代码审查为主、工具静态扫描分析为辅的静态测试方案.检查单法侧重于检查项的对照性检查与确认,变量分析法和中断访问冲突分析法侧重于变量及其变体的分析及确认,代码逻辑分析法侧重于功能级的代码整体逻辑的分析与确认,工具静态扫描分析法则以全自主研发的典型缺陷模式集为核心实现缺陷自动化检测.研究成果已在航天软件第三方评测中全面推广应用.通过对方案综合应用数据、各方法应用数据的统计分析表明,工程应用效果良好,静态测试的缺陷检测率普遍提升至95%以上.相关分析方法和分析思路对软件缺陷自动化检测工具的研发以及传统软件测试方法的持续改进均具有一定的参考和推动作用.

     

    Abstract: The operation scenarios of spacecraft on orbit are highly complex, and some scenarios are difficult to achieve dynamic verification in the real state on the ground. Through the analysis and research on the defects found in recent 30 years, a set of static testing scheme based on manual code inspection and supplemented by tool static scanning analysis is proposed, including checklist method, variable analysis method, interrupt access conflict analysis method, code logic analysis method and tool static scanning analysis method. Among them, the checklist method focuses on the comparative inspection and confirmation of inspection items, the variable analysis method and interrupted access conflict analysis method focus on the analysis and confirmation of variables and their variants, and the code logic analysis method focuses on the analysis and confirmation of the overall logic of the code at the function level, the tool static scanning analysis method takes the typical defect pattern set as the core to realize automatic defect detection. The research results have been widely used in the third-party evaluation of aerospace software. The comprehensive application data and the various methods application data show that the engineering application effect is good, and the defect detection rate of static test is generally increased to more than 95%. The relevant analysis methods and ideas have a certain reference and promotion for the research and development of software defect automatic detection tools and the continuous improvement of traditional software testing methods.

     

/

返回文章
返回