动态二进制翻译中数据预取优化研究

ynamic optimization is an important research subject in binary translation systems. As a way of memory optimization, data prefetching improves application’s performance on modern processor architectures. This paper proposes a SuperBlock based dynamic data prefetching scheme for dynamic translation, it uses software instrumentation to collect program’s load delay information and build SuperBlocks. Then the data flow analysis is performed on SuperBlocks to generate the RDUG(register define and use graph). After that, data prefetching instructions are inserted for the delinquent loads. Experiments on the Loongson DigitalBridge dynamic binary translation system shows that the prefetching scheme can achieve 2.53% improvement on average for SPEC2000 float programs while the overhead is less than 0.5%.