License: Creative Commons Attribution 3.0 Unported license (CC BY 3.0)
When quoting this document, please refer to the following
DOI: 10.4230/LIPIcs.OPODIS.2017.7
URN: urn:nbn:de:0030-drops-86275
URL: http://dagstuhl.sunsite.rwth-aachen.de/volltexte/2018/8627/
Go to the corresponding LIPIcs Volume Portal

Zhou, Tingzhe ; Luchangco, Victor ; Spear, Michael

Extending Transactional Memory with Atomic Deferral

pdf-format:
LIPIcs-OPODIS-2017-7.pdf (0.6 MB)

Abstract

This paper introduces atomic deferral, an extension to TM that allows programmers to move long-running or irrevocable operations out of a transaction while maintaining serializability: the transaction and its de- ferred operation appear to execute atomically from the perspective of other transactions. Thus, program- mers can adapt lock-based programs to exploit TM with relatively little effort and without sacrificing scalability by atomically deferring the problematic operations. We demonstrate this with several use cases for atomic deferral, as well as an in-depth analysis of its use on the PARSEC dedup benchmark, where we show that atomic deferral enables TM to be competitive with well-designed lock-based code.

BibTeX - Entry

@InProceedings{zhou_et_al:LIPIcs:2018:8627,
  author =	{Tingzhe Zhou and Victor Luchangco and Michael Spear},
  title =	{{Extending Transactional Memory with Atomic Deferral}},
  booktitle =	{21st International Conference on Principles of Distributed Systems (OPODIS 2017)},
  pages =	{7:1--7:17},
  series =	{Leibniz International Proceedings in Informatics (LIPIcs)},
  ISBN =	{978-3-95977-061-3},
  ISSN =	{1868-8969},
  year =	{2018},
  volume =	{95},
  editor =	{James Aspnes and Alysson Bessani and Pascal Felber and Jo{\~a}o Leit{\~a}o},
  publisher =	{Schloss Dagstuhl--Leibniz-Zentrum fuer Informatik},
  address =	{Dagstuhl, Germany},
  URL =		{http://drops.dagstuhl.de/opus/volltexte/2018/8627},
  URN =		{urn:nbn:de:0030-drops-86275},
  doi =		{10.4230/LIPIcs.OPODIS.2017.7},
  annote =	{Keywords: Transactional Memory, Concurrency, Synchronization, I/O}
}

Keywords: Transactional Memory, Concurrency, Synchronization, I/O
Collection: 21st International Conference on Principles of Distributed Systems (OPODIS 2017)
Issue Date: 2018
Date of publication: 28.03.2018

DROPS-Home | Fulltext Search | Imprint | Privacy Published by LZI