Optimizing memory transactions for large-scale programs

作者: Fernando Miguel Carvalho , João Cachopo

DOI: 10.1016/J.JPDC.2015.12.001

关键词: Lock (computer science)Concurrent computingKey (cryptography)Software transactional memoryObject (computer science)Identification (information)Computer scienceRecord lockingMetadataDistributed computing

摘要: Even though Software Transactional Memory (STM) is one of the most promising approaches to simplify concurrent programming, current STM implementations incur significant overheads that render them impractical for many real-sized programs. The key insight this work we do not need use same costly barriers all memory managed by a application, if only small fraction under contention-lightweight may be used in case. In work, propose new solution based on an approach adaptive object metadata (AOM) promote fast path access objects are contention. We show able make performance competitive with best fine-grained lock-based some more challenging benchmarks. Display Omitted A technique eliminates extra metadata.AOM LICM (lightweight identification captured memory) provide non-contended objects.Results rivals lock large-scale benchmark.Integrated Deuce full support in-place required and AOM.Innovative adaptation STM: maintains original API, enhances any existing STM.

参考文章(77)
Frank Yellin, Tim Lindholm, Java Virtual Machine Specification Addison-Wesley Longman Publishing Co., Inc.. ,(1999)
Anurag Acharya, Christopher Heriot, Michael F. Spear, William N. Scherer, Michael L Scott, David D. Eisenstat, Virendra Jayant Marathe, Lowering the Overhead of Software Transactional Memory ,(2006)
Fernando Miguel Carvalho, Joao Cachopo, STM with transparent API considered harmful international conference on algorithms and architectures for parallel processing. pp. 326- 337 ,(2011) , 10.1007/978-3-642-24650-0_28
Fernando Miguel Carvalho, João Cachopo, Lightweight Identification of Captured Memory for Software Transactional Memory Algorithms and Architectures for Parallel Processing. pp. 15- 29 ,(2013) , 10.1007/978-3-319-03859-9_2
D. P. Reed, NAMING AND SYNCHRONIZATION IN A DECENTRALIZED COMPUTER SYSTEM Massachusetts Institute of Technology. ,(1978)
Nir Shavit, David Dice, What Really Makes Transactions Faster ,(2006)
Torvald Riegel, Pascal Felber, Christof Fetzer, A Lazy Snapshot Algorithm with Eager Validation Lecture Notes in Computer Science. pp. 284- 298 ,(2006) , 10.1007/11864219_20