DeepRace: A learning-based data race detector

TehraniJamsaz, Ali; Khaleel, Mohammed; Akbari, Reza; Jannesari, Ali

doi:10.1109/icstw52544.2021.00046

Cited by 7 publications

(3 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Moreover, this representation is hardware specific, thus limiting the cross-architecture characterization. An alternative to avoid these issues is to apply natural language techniques on the source code to represent the applications [21]- [23]. They mainly consider source code as a sequence of tokens.…”

Section: A Graph Representation Of Programsmentioning

confidence: 99%

Learning Intermediate Representations using Graph Neural Networks for NUMA and Prefetchers Optimization

TehraniJamsaz,

Popov,

Dutta

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

There is a large space of NUMA and hardware prefetcher configurations that can significantly impact the performance of an application. Previous studies have demonstrated how a model can automatically select configurations based on the dynamic properties of the code to achieve speedups. This paper demonstrates how the static Intermediate Representation (IR) of the code can guide NUMA/prefetcher optimizations without the prohibitive cost of performance profiling. We propose a method to create a comprehensive dataset that includes a diverse set of intermediate representations along with optimum configurations. We then apply a graph neural network model in order to validate this dataset. We show that our static intermediate representation based model achieves 80% of the performance gains provided by expensive dynamic performance profiling based strategies. We further develop a hybrid model that uses both static and dynamic information. Our hybrid model achieves the same gains as the dynamic models but at a reduced cost by only profiling 30% of the programs.

show abstract

Section: A Graph Representation Of Programsmentioning

confidence: 99%

Learning Intermediate Representations using Graph Neural Networks for NUMA and Prefetchers Optimization

TehraniJamsaz,

Popov,

Dutta

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…To improve performance with many threads, it is desirable to consider alternative methods in the future, such as using atomic operations, etc. Also, promising directions are considered to be the use of special tools, for example, ROMP [8], DataRaceBench [9], and new approaches to detecting thread races based on deep neural network models [10].…”

mentioning

confidence: 99%

Дослідження Проблем Синхронізації Та Захисту Даних При Реалізації Багатопоточних Програм

Oleg,

Inna,

Victoria

et al. 2024

View full text Add to dashboard Cite

The issue of shared data usage by threads is especially relevant in modern multi-core and multiprocessor systems. The main problems of implementing multithreaded programs are race conditions, deadlocks, and thread starvation. The aim of the work is to solve the problem of thread racing in multithreaded calculations of resource-intensive tasks with parallel access to shared data using appropriate synchronization mechanisms, such as mutexes. A multithreaded algorithm for implementing a typical task of processing large data arrays with protection of the critical area in concurrent programs running on multiprocessor and multi-core systems has been developed and researched.

show abstract

“…[35] works on applying LSTM for inferring the code a specific type of missing error as variable declarations. [21] proposes DeepRace, a bug finding tool for data race errors in parallel programs using advantages of Convolution Neural Network (CNN). We can think about integrating the new models from these works to NMT to increase the performance of NMT in SE.…”

Section: Rq5: Analysis On the Accuracy Of Prefixmap On Ambiguous Tokensmentioning

confidence: 99%

A study on the performance of statistical machine translation and neural machine translation for prefix resolution in software engineering

Phan

View full text Add to dashboard Cite

show abstract

DeepRace: A learning-based data race detector

Cited by 7 publications

References 35 publications

Learning Intermediate Representations using Graph Neural Networks for NUMA and Prefetchers Optimization

Learning Intermediate Representations using Graph Neural Networks for NUMA and Prefetchers Optimization

Дослідження Проблем Синхронізації Та Захисту Даних При Реалізації Багатопоточних Програм

A study on the performance of statistical machine translation and neural machine translation for prefix resolution in software engineering

Contact Info

Product

Resources

About