基于Trace-Cache的多级动态优化框架设计

Trace-Cache Based Framework for Multi-Level Dynamic Optimization

对指令集进行扩展和添加新功能部件是提高处理器性能的有效途径.为了充分利用新的体系结构扩展,已有应用必需经过全新的优化编译.对于跨体系结构优化而言,二进制翻译已经被证明是一种行之有效的技术.本文结合trace技术和动态二进制翻译优化技术,提出一种多级动态优化框架结构,无需静态重新优化编译,在程序动态运行期间,引入多级动态优化方法和扩展指令调度.模拟结果显示该结构具有能有效形成大尺寸的指令调度窗口,准确选择热点代码及优化方法,有效提升旧有应用性能的优点,并有实现灵活,可扩展好等特点.

New extension of instruction sot and new add-on function unit can improve the performance of microprocessor greatly. All the applications should be recompiled and rebuilt, otherwise they can＇ t benefit from those new instruction. This paper proposes a framework of multi-level dynamic optimization, which introduces instruction scheduling and optimizing for the architecture extension based on trace-cache in rantime. Experimental results show that it can enlarge the instruction window to select hot cedes and schedul- ing methods efficiently and effectively, and leverage the performance of original application without the need of recompiling. In addition,this framework is flexible and scalable to new optimizing chance and various platforms.