| 基于BWDSP的字符串与内存处理函数优化 |
| |
| 引用本文： | 张仁高,郑启龙,王向前.基于BWDSP的字符串与内存处理函数优化[J].计算机系统应用,2017,26(7):167-172. |
| |
| 作者姓名： | 张仁高 郑启龙 王向前 |
| |
| 作者单位： | 中国科学技术大学 计算机科学与技术学院, 合肥 230027,中国科学技术大学 计算机科学与技术学院, 合肥 230027,中国电子科技集团公司第三十八研究所, 合肥 230088 |
| |
| 基金项目： | 国家核高基重大专项（2012ZX01034001-001） |
| |
| 摘 要： | 面向BWDSP的体系结构分析了字符串与内存处理函数汇编优化方法，基于向量化与软件流水的优化技术，通过利用高效访存指令、能够提升循环执行效率的零开销循环机制、指令重排技术，结合具体功能函数的循环特性，展开针对字符串与内存处理函数的指令级并行性挖掘.实验结果表明，这些库函数的优化效率能够达到硬件平台提供函数性能理论运行时间的1.5倍以下，对BWDSP平台整体性能提升具有重要意义.
|
| 关 键 词： | 字符串与内存处理函数 BWDSP 函数优化 向量化与软件流水 特殊指令 并行性 |
| 收稿时间： | 2016/11/19 0:00:00 |
| 修稿时间： | 2016/11/24 0:00:00 |
Optimization of String and Memory Functions Based on BWDSP |
| |
| ZHANG Ren-Gao,ZHENG Qi-Long and WANG Xiang-Qian.Optimization of String and Memory Functions Based on BWDSP[J].Computer Systems& Applications,2017,26(7):167-172. |
| |
| Authors: | ZHANG Ren-Gao ZHENG Qi-Long and WANG Xiang-Qian |
| |
| Affiliation: | School of Computer Science and Technology, University of Science and Technology of China, Hefei 230027, China,School of Computer Science and Technology, University of Science and Technology of China, Hefei 230027, China and No.38 Research Institute, China Electronics Technology Group Corporation, Hefei 230088, China |
| |
| Abstract: | For architecture of BWDSP, this paper analyzes the implementation of the optimizations, based on vectorization and software pipelining of parallel optimization technique, including special memory access instruction, zero-overhead looping instruction of improving efficiency in the loop, Instruction-level Parallelism(ILP), combined with the specific function of loop characteristics, expansion for the string and memory function of instruction level parallel mining. The experimental results show that the optimization rates of most functions of the theory have a running time of 1.5 times on hardware platform, which is of great importance to enhance the platform performance of BWDSP. |
| |
| Keywords: | string and memory functions BWDSP optimization of function vectorization and software pipelining special instruction parallelism |
|
| 点击此处可从《计算机系统应用》浏览原始摘要信息 |
|
点击此处可从《计算机系统应用》下载全文 |
|
