面向X10的数据流程序编译优化方法研究-计算机应用技术专业毕业论文.docxVIP

华 华 中 科 技 大 学 硕 士 学 位 论 文 II II Abstract As the traditional single-core processor has been influenced by the power consumption, leakage and electromagnetic interference, its performance improvement has become a bottleneck.In order to improve the overall performance of the processors, multi-core processors have gradually replaced it, kept working frequency, reduced the average power consumption of the processor, and obtained higher energy efficiency. For another, in order to utilize the resources of the underlying hardware interface explicitly, the parallel programming model for parallel computing on multicore platforms otherwise increases programming complexity. However, the emergence of stream programming model solved such problems, which not only takes full advantage of the underlying resources, but also provides simple, efficient, scalable programming method. For the difficulty of the poor programming efficiency and execution efficiency of current stream programming model, we designed and implemented a language COStream and the corresponding optimized compiler system. The COStream language bases on the modifications and extensions of the standard C language syntax, supporting for the operation of the grammatical structure: composite, operator, stream. After the phase of lexical analysis, syntax analysis, semantic analysis, the complier translates the languagege code into an intermediate syntax tree and establish synchronous data flow diagram. Combined with the characteristics of the IBM parallel programming language X10, the compiler established three optimization strategies: generating optimized code for the X10 language, optimizing the division of SDF graph, optimizing the communications for the underlying hardware resources. The Code generation optimization reduced the amount of X10 code; the task division optimization obtained load balancing, and avoided deadlock, at the same time reducd communication overhead; the communication optimization disti