Ying Huan, Miao Si-Wei, Han Li-fang, Yu Wen-Hao, Zhou Liang
{"title":"基于依赖对的快速确定性重放方法","authors":"Ying Huan, Miao Si-Wei, Han Li-fang, Yu Wen-Hao, Zhou Liang","doi":"10.1109/ICIST.2018.8426126","DOIUrl":null,"url":null,"abstract":"Deterministic replay plays an important role for parallel program debugging. Related work focused on how to efficiently record undeterminstic events in parallel programs, while ignored poor performance in replay-runs. This paper proposes a quick deterministic replay method based on dependence pair to acquire excellent performance in replay-runs. It conducts offline analysis for record log file to recognize the record items with dependency relationship based on topological graph, and each thread only executes these record items in order in replay-runs. Performance evaluation shows that this method gets better performance in replay-runs than traditional serial replay method.","PeriodicalId":331555,"journal":{"name":"2018 Eighth International Conference on Information Science and Technology (ICIST)","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2018-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"A Quick Deterministic Replay Method Based on Dependence Pair\",\"authors\":\"Ying Huan, Miao Si-Wei, Han Li-fang, Yu Wen-Hao, Zhou Liang\",\"doi\":\"10.1109/ICIST.2018.8426126\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Deterministic replay plays an important role for parallel program debugging. Related work focused on how to efficiently record undeterminstic events in parallel programs, while ignored poor performance in replay-runs. This paper proposes a quick deterministic replay method based on dependence pair to acquire excellent performance in replay-runs. It conducts offline analysis for record log file to recognize the record items with dependency relationship based on topological graph, and each thread only executes these record items in order in replay-runs. Performance evaluation shows that this method gets better performance in replay-runs than traditional serial replay method.\",\"PeriodicalId\":331555,\"journal\":{\"name\":\"2018 Eighth International Conference on Information Science and Technology (ICIST)\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2018-06-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2018 Eighth International Conference on Information Science and Technology (ICIST)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/ICIST.2018.8426126\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2018 Eighth International Conference on Information Science and Technology (ICIST)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICIST.2018.8426126","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
A Quick Deterministic Replay Method Based on Dependence Pair
Deterministic replay plays an important role for parallel program debugging. Related work focused on how to efficiently record undeterminstic events in parallel programs, while ignored poor performance in replay-runs. This paper proposes a quick deterministic replay method based on dependence pair to acquire excellent performance in replay-runs. It conducts offline analysis for record log file to recognize the record items with dependency relationship based on topological graph, and each thread only executes these record items in order in replay-runs. Performance evaluation shows that this method gets better performance in replay-runs than traditional serial replay method.