Dynamic Cache Contention Detection in Multi-threaded Applications
In today's multi-core systems, cache contention due to true and false sharing can cause unexpected and significant performance degradation. A detailed understanding of a given multi-threaded application's behavior is required to precisely identify such performance bottlenecks. Traditionall...
| Päätekijät: | , , , , , |
|---|---|
| Muut tekijät: | |
| Aineistotyyppi: | Artikkeli |
| Kieli: | en_US |
| Julkaistu: |
Association for Computing Machinery / ACM Special Interest Group on Programming Languages./ ACM Special Interest Group in Operating Systems.
2011
|
| Linkit: | http://hdl.handle.net/1721.1/62586 https://orcid.org/0000-0002-7231-7643 |
| _version_ | 1826206816722223104 |
|---|---|
| author | Zhao, Qin Koh, David F. Raza, Syed A. Amarasinghe, Saman P. Bruening, Derek Wong, Weng-Fai |
| author2 | Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory |
| author_facet | Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory Zhao, Qin Koh, David F. Raza, Syed A. Amarasinghe, Saman P. Bruening, Derek Wong, Weng-Fai |
| author_sort | Zhao, Qin |
| collection | MIT |
| description | In today's multi-core systems, cache contention due to true and false sharing can cause unexpected and significant performance degradation. A detailed understanding of a given multi-threaded application's behavior is required to precisely identify such performance bottlenecks. Traditionally, however, such diagnostic information can only be obtained after lengthy simulation of the memory hierarchy.
In this paper, we present a novel approach that efficiently analyzes interactions between threads to determine thread correlation and detect true and false sharing. It is based on the following key insight: although the slowdown caused by cache contention depends on factors including the thread-to-core binding and parameters of the memory hierarchy, the amount of data sharing is primarily a function of the cache line size and application behavior. Using memory shadowing and dynamic instrumentation, we implemented a tool that obtains detailed sharing information between threads without simulating the full complexity of the memory hierarchy. The runtime overhead of our approach --- a 5x slowdown on average relative to native execution --- is significantly less than that of detailed cache simulation. The information collected allows programmers to identify the degree of cache contention in an application, the correlation among its threads, and the sources of significant false sharing. Using our approach, we were able to improve the performance of some applications up to a factor of 12x. For other contention-intensive applications, we were able to shed light on the obstacles that prevent their performance from scaling to many cores. |
| first_indexed | 2024-09-23T13:39:01Z |
| format | Article |
| id | mit-1721.1/62586 |
| institution | Massachusetts Institute of Technology |
| language | en_US |
| last_indexed | 2024-09-23T13:39:01Z |
| publishDate | 2011 |
| publisher | Association for Computing Machinery / ACM Special Interest Group on Programming Languages./ ACM Special Interest Group in Operating Systems. |
| record_format | dspace |
| spelling | mit-1721.1/625862022-09-28T15:16:51Z Dynamic Cache Contention Detection in Multi-threaded Applications Zhao, Qin Koh, David F. Raza, Syed A. Amarasinghe, Saman P. Bruening, Derek Wong, Weng-Fai Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science Amarasinghe, Saman P. Zhao, Qin Koh, David F. Raza, Syed A. Amarasinghe, Saman P. In today's multi-core systems, cache contention due to true and false sharing can cause unexpected and significant performance degradation. A detailed understanding of a given multi-threaded application's behavior is required to precisely identify such performance bottlenecks. Traditionally, however, such diagnostic information can only be obtained after lengthy simulation of the memory hierarchy. In this paper, we present a novel approach that efficiently analyzes interactions between threads to determine thread correlation and detect true and false sharing. It is based on the following key insight: although the slowdown caused by cache contention depends on factors including the thread-to-core binding and parameters of the memory hierarchy, the amount of data sharing is primarily a function of the cache line size and application behavior. Using memory shadowing and dynamic instrumentation, we implemented a tool that obtains detailed sharing information between threads without simulating the full complexity of the memory hierarchy. The runtime overhead of our approach --- a 5x slowdown on average relative to native execution --- is significantly less than that of detailed cache simulation. The information collected allows programmers to identify the degree of cache contention in an application, the correlation among its threads, and the sources of significant false sharing. Using our approach, we were able to improve the performance of some applications up to a factor of 12x. For other contention-intensive applications, we were able to shed light on the obstacles that prevent their performance from scaling to many cores. 2011-05-04T19:18:08Z 2011-05-04T19:18:08Z 2011-03 Article http://purl.org/eprint/type/ConferencePaper 978-1-4503-0687-4 http://hdl.handle.net/1721.1/62586 Zhao, Qin et al. “Dynamic Cache Contention Detection in Multi-threaded Applications.” Proceedings of the 7th ACM SIGPLAN/SIGOPS International Conference on Virtual Execution Environments - VEE ’11. Newport Beach, California, USA, 2011. 27. Copyright c2011 ACM https://orcid.org/0000-0002-7231-7643 en_US http://dx.doi.org/10.1145/1952682.1952688 VEE Proceedings (ACM SIGPLAN SIGOPS International Conference on Virtual Execution Environments) Creative Commons Attribution-Noncommercial-Share Alike 3.0 http://creativecommons.org/licenses/by-nc-sa/3.0/ application/pdf Association for Computing Machinery / ACM Special Interest Group on Programming Languages./ ACM Special Interest Group in Operating Systems. MIT web domain |
| spellingShingle | Zhao, Qin Koh, David F. Raza, Syed A. Amarasinghe, Saman P. Bruening, Derek Wong, Weng-Fai Dynamic Cache Contention Detection in Multi-threaded Applications |
| title | Dynamic Cache Contention Detection in Multi-threaded Applications |
| title_full | Dynamic Cache Contention Detection in Multi-threaded Applications |
| title_fullStr | Dynamic Cache Contention Detection in Multi-threaded Applications |
| title_full_unstemmed | Dynamic Cache Contention Detection in Multi-threaded Applications |
| title_short | Dynamic Cache Contention Detection in Multi-threaded Applications |
| title_sort | dynamic cache contention detection in multi threaded applications |
| url | http://hdl.handle.net/1721.1/62586 https://orcid.org/0000-0002-7231-7643 |
| work_keys_str_mv | AT zhaoqin dynamiccachecontentiondetectioninmultithreadedapplications AT kohdavidf dynamiccachecontentiondetectioninmultithreadedapplications AT razasyeda dynamiccachecontentiondetectioninmultithreadedapplications AT amarasinghesamanp dynamiccachecontentiondetectioninmultithreadedapplications AT brueningderek dynamiccachecontentiondetectioninmultithreadedapplications AT wongwengfai dynamiccachecontentiondetectioninmultithreadedapplications |