代码选择在编译器的代码产生阶段是一个十分重要的任务,它的目标就是在与机器无关的中间表示代码和与处理器相关的机器指令之间寻找一种高效的映射方法。为了支持DSP处理器的SIMD指令,在传统的基于数据流树中间表示的代码选择算法的基础上,提出一种基于数据流图(DFG)的代码选择技术,它能在最大限度地挖掘和利用SIMD指令的基础上寻求对整个DFG的最优覆盖。
Code selection is an important task in code generation, where the goal is to find an efficient mapping of machine-independent intermediate code to processor-specific machine instructions, In order to support SIMD of DSP, the traditional approach has been expanded and data-flow graph-based code selection techniques have been presented, which can maximize the use of SIMD and find an optimal covering across the entire DFG.