面向RISC-V指令集架构处理器的代码压缩技术  被引量:2

Code compression technology for processors based on RISC-V instruction set architecture

在线阅读下载全文

作  者:程战涛 梁峰[1] 张国和[1] CHENG Zhan-tao;LIANG Feng;ZHANG Guo-he(School of Microelectronics,Xi'an Jiaotong University,Xi'an 710049,Shaanxi China)

机构地区:[1]西安交通大学微电子学院,陕西西安710049

出  处:《微电子学与计算机》2021年第6期13-19,共7页Microelectronics & Computer

基  金:国家自然科学基金资助项目(61474093);陕西省自然科学基金资助项目(2020JM-006);航空科学基金项目(20184370012)。

摘  要:针对嵌入式系统处理器代码量指数式增长带来的设计与验证难题,提出一种适用于RISC-V指令集架构处理器的Bitmask代码压缩技术.结合RISC-V指令集特点,设计了混合编码与分离字典相结合的Bitmask代码压缩算法;在不影响处理器结构和Cache工作机制的基础上,设计精简的硬件解压缩单元,减小了嵌入式系统处理器所需的程序内存空间.采用面向RISC-V指令集的混合编码压缩指令格式,减小原始指令码的码字长度,改善了代码压缩率;采用分离的两个字典结构,在不影响代码压缩率的前提下,减小了硬件解压缩延迟.结果表明,在RISC-V指令集架构上不增加过多硬件开销的情况下,代码压缩率平均为61.1%,大大减小了处理器所需的程序内存空间.Aiming at design and verification problems caused by the exponential growth of the code size in the embedded system processors,the Bitmask-based code compression technology for processors based on RISC-V instruction set architecture is presented.Based on the features of RISC-V instruction set,a Bitmask-based code compression algorithm combined with mixed encoding and separate dictionary is designed.Without affecting the processor structure and Cache working mechanism,a simple and efficient hardware decompression unit is designed to reduce the program memory space required by the embedded system processor.Mixed encoding format for RISC-V instruction set is used to reduce the length of the original instruction,which improves the code compression rate.Besides,two separated dictionaries are used to obtain a small hardware decompression delay without affecting the code compression rate.The experimental results show that the code compression ratio is 61.1%on average without much hardware overhead on the RISC-V instruction set architecture,which greatly reduces the program memory space required by the processor.

关 键 词:嵌入式系统处理器 RISC-V指令集 代码压缩技术 硬件解压缩单元 

分 类 号:TP332[自动化与计算机技术—计算机系统结构]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象