作者: Manuel Carro Liñares , Amadeo Casas , Manuel V. Hermenegildo
DOI:
关键词:
摘要: Most implementations of parallel logic programming rely on complex low-level machinery which is arguably difflcult to implement and modify. We explore an alternative approach aimed at taming that complexity by raising core parts the implementation source language level for particular case and-parallelism. Therefore, we handle a signiflcant portion mechanism Prolog with help comparatively small number concurrency-related primitives take care lower-level tasks such as locking, thread management, stack set etc. The does not eliminate altogether modiflcations abstract machine, but it greatly simplify them also facilitates experimenting different alternatives. show how this allows implementing both restricted unrestricted (i.e., non fork-join) parallelism. Preliminary experiments amount performance sacriflced reasonable, although granularity control required in some cases. Also, observe availability parallelism contributes better observed speedups.