A Practical Solution to the Cactus Stack Problem.
SPAA(2016)
摘要
ABSTRACTWork-stealing is a popular method for load-balancing dynamic multithreaded computations on shared-memory systems. In theory, a randomized work-stealing scheduler can achieve near linear speedup when the computation has sufficient parallelism and requires stack space that is linear in the number of processors. In practice, however, work-stealing runtimes sacrifice interoperability with serial code to achieve these bounds. For example, both Cilk and Cilk++ prohibit a C function from calling aCilk function. Other work-stealing runtime systems that do not have this restriction either lack a strong time bound, which might cause them to deliver little or no speedup in the worst case, or lack a strong space bound, which might lead to an excessive memory footprint. This problem was previously described as the cactus stack problem. In this paper, we present Fibril, a new multithreading library that supports a fork-join programming model using work-stealing. Fibril solves the cactus stack problem by (1) implementing on a cactus stack that conforms to the calling conventions of serial code and (2) returning unused memory pages of suspended stacks to the operating system to bound consumption of physical memory. Theoretically, Fibril achieves strong bounds on both time and memory usage without sacrificing interoperability with serial code. Empirically, Fibril achieves up to 3x the performance of Intel Cilk Plus and up to 8x the performance of Intel Threading Building Blocks for the 12 benchmarks we evaluated.
更多查看译文
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络