메뉴 건너뛰기




Volumn 33, Issue 1, 2008, Pages 67-69

Use lock hierarchies to avoid deadlock

Author keywords

[No Author keywords available]

Indexed keywords


EID: 38049183001     PISSN: 1044789X     EISSN: None     Source Type: Trade Journal    
DOI: None     Document Type: Article
Times cited : (3)

References (1)
  • 1
    • 38049157750 scopus 로고    scopus 로고
    • Or otherwise gets the same effect. For example, it is possible to just start trying to acquire the locks in some randomly selected order using try lock operations, and if we can't acquire them all just back-off (unlock the ones already acquired) and try a different order until we find one that works. Surprisingly, this can be more efficient than taking the locks in a hard-coded global order, although any back-off-and-retry strategy has to take care that it doesn't end up prone to livelock problems instead. But we can leave all this to the implementer; the key is that the programmer simply writes lock( /* whatever */ ) and is insulated from the details of determining the best way to keep the order consistent
    • Or otherwise gets the same effect. For example, it is possible to just start trying to acquire the locks in some randomly selected order using try lock operations, and if we can't acquire them all just back-off (unlock the ones already acquired) and try a different order until we find one that works. Surprisingly, this can be more efficient than taking the locks in a hard-coded global order, although any back-off-and-retry strategy has to take care that it doesn't end up prone to livelock problems instead. But we can leave all this to the implementer; the key is that the programmer simply writes lock( /* whatever */ ) and is insulated from the details of determining the best way to keep the order consistent


* 이 정보는 Elsevier사의 SCOPUS DB에서 KISTI가 분석하여 추출한 것입니다.