摘要
针对使用路径覆盖率作为验证目标时出现的路径数量庞大、覆盖率难以快速提高等问题,使用控制流图表示代码逻辑结构,采用数据流分析技术对待验证路径进行合理的化简.所产生的路径集合规模小、针对性强,有利于尽早发现设计错误.最后使用遗传算法生成模拟矢量,并基于北京大学“众志-863系统”芯片中的功能模块给出实验结果,说明该覆盖率模型的有效性.
To cope with problems such as the number of path is huge and path coverage can hardly increase rapidly. In this paper, we use control flow graph to describe the structure of hardware design language design and use techniques of data flow analysis to reduce the number of path. A genetic algorithm is used to generate stimuli and calculate the path coverage based on critical signals. Finally, we perform experiments on the functional modules of Unity-863 SoC. Experimental results show efficiency of this metric to detect design errors.
出处
《计算机辅助设计与图形学学报》
EI
CSCD
北大核心
2006年第8期1085-1091,共7页
Journal of Computer-Aided Design & Computer Graphics
基金
国家"八六三"高技术研究发展计划(2004AA1Z1010)
关键词
覆盖率模型
路径覆盖率
数据流分析
模拟矢量生成
遗传算法
coverage metric
path coverage
data flow analysis
simulation vector generation
genetic algorithm
