Effective Transactional Memory Execution Management for Improved Concurrency

This article describes a transactional memory execution model intended to exploit maximum parallelism from sequential and multithreaded programs. A program code section is partitioned into chunks that will be mapped onto threads and executed transactionally. These transactions run concurrently and out of order, trying to exploit maximum parallelism but managed by a specific fully distributed commit control to meet data dependencies. To accomplish correct parallel execution, a partial precedence order relation is derived from the program code section and/or defined by the programmer. When a conflict between chunks is eagerly detected, the precedence order relation is used to determine the best policy to solve the conflict that preserves the precedence order while maximizing concurrency. The model defines a new transactional state called executed but not committed. This state allows exploiting concurrency on two levels: intrathread and interthread. Intrathread concurrency is improved by having pending uncommitted transactions while executing a new one in the same thread. The new state improves interthread concurrency because it permits out-of-order transaction commits regarding the precedence order. Our model has been implemented in a lightweight software transactional memory system, TinySTM, and has been evaluated on a set of benchmarks obtaining an important performance improvement over the baseline TM system.

[1]  Dean M. Tullsen,et al.  Mapping Out a Path from Hardware Transactional Memory to Speculative Multithreading , 2009, 2009 18th International Conference on Parallel Architectures and Compilation Techniques.

[2]  Sergio Pissanetzky,et al.  Sparse Matrix Technology , 1984 .

[3]  Samuel Williams,et al.  Roofline: an insightful visual performance model for multicore architectures , 2009, CACM.

[4]  Christopher J. Hughes,et al.  Performance evaluation of Intel® Transactional Synchronization Extensions for high-performance computing , 2013, 2013 SC - International Conference for High Performance Computing, Networking, Storage and Analysis (SC).

[5]  Kunle Olukotun,et al.  The OpenTM Transactional Application Programming Interface , 2007, 16th International Conference on Parallel Architecture and Compilation Techniques (PACT 2007).

[6]  Martin Schulz,et al.  What scientific applications can benefit from hardware transactional memory? , 2012, 2012 International Conference for High Performance Computing, Networking, Storage and Analysis.

[7]  Seon Wook Kim,et al.  Runtime parallelization of legacy code on a transactional memory system , 2011, HiPEAC.

[8]  Josep Torrellas,et al.  BulkSMT: Designing SMT processors for atomic-block execution , 2012, IEEE International Symposium on High-Performance Comp Architecture.

[9]  Scott A. Mahlke,et al.  Parallelizing sequential applications on commodity hardware using a low-cost software transactional memory , 2009, PLDI '09.

[10]  Binoy Ravindran,et al.  HydraVM: extracting parallelism from legacy sequential code using STM , 2012, HotPar'12.

[11]  Maurice Herlihy,et al.  Wait-free synchronization , 1991, TOPL.

[12]  Emmett Witchel,et al.  Dependence-aware transactional memory for increased concurrency , 2008, 2008 41st IEEE/ACM International Symposium on Microarchitecture.

[13]  Barna L. Bihari.  Transactional Memory for Unstructured Mesh Simulations , 2013, J. Sci. Comput..

[14]  Kunle Olukotun,et al.  STAMP: Stanford Transactional Applications for Multi-Processing , 2008, 2008 IEEE International Symposium on Workload Characterization.

[15]  Torvald Riegel,et al.  Time-based transactional memory with scalable time bases , 2007, SPAA '07.

[16]  S. Pissanetzky.  Sparse Matrix Algebra , 1984 .

[17]  Kai Li,et al.  The PARSEC benchmark suite: Characterization and architectural implications , 2008, 2008 International Conference on Parallel Architectures and Compilation Techniques (PACT).

[18]  Emilio L. Zapata,et al.  A compiler method for the parallel execution of irregular reductions in scalable shared memory multiprocessors , 2000, ICS '00.

[19]  D. Geer,et al.  Chip makers turn to multicore processors , 2005, Computer.

[20]  Luca Benini,et al.  SoC-TM: Integrated HW/SW support for transactional memory programming on embedded MPSoCs , 2011, 2011 Proceedings of the Ninth IEEE/ACM/IFIP International Conference on Hardware/Software Codesign and System Synthesis (CODES+ISSS).

[21]  James R. Larus,et al.  Transactional Memory, 2nd edition , 2010, Transactional Memory.

[22]  Torvald Riegel,et al.  Dynamic performance tuning of word-based software transactional memory , 2008, PPoPP.

[23]  Larry Rudolph,et al.  Basic Techniques for the Efficient Coordination of Very Large Numbers of Cooperating Sequential Processors , 1983, TOPL.

[24]  Rachid Guerraoui,et al.  Unifying Thread-Level Speculation and Transactional Memory , 2012, Middleware.

[25]  Kunle Olukotun,et al.  Transactional memory coherence and consistency , 2004, Proceedings. 31st Annual International Symposium on Computer Architecture, 2004..

[26]  Håkan Grahn,et al.  Transactional memory , 2010, J. Parallel Distributed Comput..

[27]  Richard F. Barrett,et al.  Matrix Market: a web resource for test matrix collections , 1996, Quality of Numerical Software.

[28]  Maurice Herlihy,et al.  Transactional Memory: Architectural Support For Lock-free Data Structures , 1993, Proceedings of the 20th Annual International Symposium on Computer Architecture.

[29]  Mikel Luján,et al.  Advanced Concurrency Control for Transactional Memory Using Transaction Commit Rate , 2008, Euro-Par.

[30]  Tarek S. Abdelrahman,et al.  Hardware Support for Relaxed Concurrency Control in Transactional Memory , 2010, 2010 43rd Annual IEEE/ACM International Symposium on Microarchitecture.

[31]  Gurindar S. Sohi,et al.  Multiscalar processors , 1995, Proceedings 22nd Annual International Symposium on Computer Architecture.

[32]  Helmar Burkhart,et al.  Automatic code generation and tuning for stencil kernels on modern shared memory architectures , 2011, Computer Science - Research and Development.