Detection of false sharing using machine learning

Jayasena, Sanath; Amarasinghe, Saman; Abeyweera, Asanka; Amarasinghe, Gayashan; Silva, Himeshi De; Rathnayake, Sunimal; Meng, Xiaofeng; Liu, Yanbin

doi:10.1145/2503210.2503269

Cited by 25 publications

(17 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Many projects focus on detecting one type of contention. Collecting hardware counters has been extensively used for detecting cache contention [2], [4], [43], [44], [45], [46], NUMA effects [47], or false sharing [48]. Several works simulate the memory subsystem to detect cache issues [6], or to detect false-sharing [28], [49].…”

Section: Related Workmentioning

confidence: 99%

Using Differential Execution Analysis to Identify Thread Interference

Bouksiaa

Lescouet

Voron

et al. 2019

IEEE Trans. Parallel Distrib. Syst.

View full text Add to dashboard Cite

Understanding the performance of a multi-threaded application is difficult. The threads interfere when they access the same shared resource, which slows down their execution. Unfortunately, current profiling tools report the hardware components or the synchronization primitives that saturate, but they cannot tell if the saturation is the cause of a performance bottleneck. In this paper, we propose a holistic metric able to pinpoint the blocks of code that suffer interference the most, regardless of the interference cause. Our metric uses performance variation as a universal indicator of interference problems. With an evaluation of 27 applications we show that our metric can identify interference problems caused by 6 different kinds of interference in 9 applications. We are able to easily remove 7 of the bottlenecks, which leads to a performance improvement of up to 9 times.

show abstract

Section: Related Workmentioning

confidence: 99%

Using Differential Execution Analysis to Identify Thread Interference

Bouksiaa

Lescouet

Voron

et al. 2019

IEEE Trans. Parallel Distrib. Syst.

View full text Add to dashboard Cite

show abstract

“…As prior work [16,19,20,32], we perform experiments on two well-known benchmark suites: Phoenix [27] and PARSEC [3]. We intentionally use 16 threads in order to run applications for sufficiently long time, as Cheetah needs enough samples to detect false sharing problems.…”

Section: Evaluated Applicationsmentioning

confidence: 99%

“…All existing PMU-based approaches actually detect false sharing by monitoring cache related events. Jayasena et al use the machine learning approach to derive the potential pattern of false sharing by monitoring cache misses, TLB events, interactions among cores, and resources stalls [16]. DARWIN collects cache coherence events during the first round, then identifies possible memory accesses on data structures with frequent cache invalidations during the second round [30].…”

Section: Os-related Approaches Sheriff Proposes Turning Threadsmentioning

confidence: 99%

“…Existing false sharing detectors fall short in several ways. First, most tools cannot distinguish true and false sharing, or require substantial manual effort to identify optimization opportunities [9,12,13,16,18,24,26,30,31]. Second, tools [9,18,20,28,32] based on memory access instrumentation introduce high runtime overhead, hindering their applicability to real, longrun applications.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Cheetah: detecting false sharing efficiently and effectively

Liu

2016

Proceedings of the 2016 International Symposium on Code Generation and Optimization

View full text Add to dashboard Cite

False sharing is a notorious performance problem that may occur in multithreaded programs when they are running on ubiquitous multicore hardware. It can dramatically degrade the performance by up to an order of magnitude, significantly hurting the scalability. Identifying false sharing in complex programs is challenging. Existing tools either incur significant performance overhead or do not provide adequate information to guide code optimization. To address these problems, we develop Cheetah, a profiler that detects false sharing both efficiently and effectively. Cheetah leverages the lightweight hardware performance monitoring units (PMUs) that are available in most modern CPU architectures to sample memory accesses. Cheetah develops the first approach to quantify the optimization potential of false sharing instances without actual fixes, based on the latency information collected by PMUs. Cheetah precisely reports false sharing and provides insightful optimization guidance for programmers, while adding less than 7% runtime overhead on average. Cheetah is ready for real deployment.

show abstract

“…Sanath et al designed a machine learning based approach to detect false sharing problems. They train their classifier on mini-programs and apply this classifier to general programs [13]. Instead of instrumenting memory accesses, this tool relies on hardware performance counters to collect memory accesses events.…”

Section: False Sharing Detectionmentioning

confidence: 99%

Predator

Liu

Tian

et al. 2014

Proceedings of the 19th ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming

View full text Add to dashboard Cite

False sharing is a notorious problem for multithreaded applications that can drastically degrade both performance and scalability. Existing approaches can precisely identify the sources of false sharing, but only report false sharing actually observed during execution; they do not generalize across executions. Because false sharing is extremely sensitive to object layout, these detectors can easily miss false sharing problems that can arise due to slight differences in memory allocation order or object placement decisions by the compiler. In addition, they cannot predict the impact of false sharing on hardware with different cache line sizes. This paper presents PREDATOR, a predictive softwarebased false sharing detector. PREDATOR generalizes from a single execution to precisely predict false sharing that is latent in the current execution. PREDATOR tracks accesses within a range that could lead to false sharing given different object placement. It also tracks accesses within virtual cache lines, contiguous memory ranges that span actual hardware cache lines, to predict sharing on hardware platforms with larger cache line sizes. For each, it reports the exact program location of predicted false sharing problems, ranked by their projected impact on performance. We evaluate PREDA-TOR across a range of benchmarks and actual applications. PREDATOR identifies problems undetectable with previous tools, including two previously-unknown false sharing problems, with no false positives. PREDATOR is able to immediately locate false sharing problems in MySQL and the Boost library that had eluded detection for years.

show abstract

Detection of false sharing using machine learning

Cited by 25 publications

References 26 publications

Using Differential Execution Analysis to Identify Thread Interference

Using Differential Execution Analysis to Identify Thread Interference

Cheetah: detecting false sharing efficiently and effectively

Predator

Contact Info

Product

Resources

About