Maximal sound predictive race detection with control flow abstraction

Huang, Jeff; Meredith, Patrick; Roşu, Grigore

doi:10.1145/2594291.2594315

Cited by 142 publications

(143 citation statements)

References 29 publications

Supporting

Mentioning

143

Contrasting

Order By: Relevance

“…A path through the computation lattice is one possible global trace of the system and it is typical to consider all such traces. In the context of multithreaded programs there exist methods generalising the happensbefore relationship by considering additional causal relationships [38,65]. This can lead to more possible global traces being explored, increasing the chances of finding buggy behaviour even if this behaviour was not observed at runtime.…”

Section: Discussionmentioning

confidence: 99%

What Is a Trace? A Runtime Verification Perspective

Reger

Havelund

2016

Leveraging Applications of Formal Methods, Verification and Validation: Discussion, Dissemination, Applications

View full text Add to dashboard Cite

Abstract. Runtime Monitoring or Verification deals with traces. In its most simple form a monitoring system takes a trace produced by a system and a specification of correct behaviour and checks if the trace conforms to the specification. More complex applications may introduce notions of feedback and reaction. The notion that unifies the field is that we can abstract the runtime behaviour of a system by an execution trace and check this for conformance. However, there is little uniform understanding of what a trace is. This is most keenly seen when comparing theoretical and practical work. This paper surveys the different notions of trace and reflects on the related issues.

show abstract

Section: Discussionmentioning

confidence: 99%

What Is a Trace? A Runtime Verification Perspective

Reger

Havelund

2016

Leveraging Applications of Formal Methods, Verification and Validation: Discussion, Dissemination, Applications

View full text Add to dashboard Cite

show abstract

“…A wide spectrum of static and dynamic techniques have been proposed [2,11,[13][14][15]28] to combat races. Static techniques [2,13,28] must be conservative and therefore report false-positives, while dynamic techniques [11,14,15,24] miss races which do not occur in the observed execution and introduce high overheads.…”

Section: Related Workmentioning

confidence: 99%

“…Static techniques [2,13,28] must be conservative and therefore report false-positives, while dynamic techniques [11,14,15,24] miss races which do not occur in the observed execution and introduce high overheads. Acculock [40] is a hybrid lockset -happens-before data race detector, balancing precision and coverage by exploring thread interleavings which do not occur in the observed execution.…”

Section: Related Workmentioning

confidence: 99%

Automatic detection of extended data-race-free regions

Jimborean

Waern

Ekemark

et al. 2017

2017 IEEE/ACM International Symposium on Code Generation and Optimization (CGO)

View full text Add to dashboard Cite

Data-race-free (DRF) parallel programming becomes a standard as newly adopted memory models of mainstream programming languages such as C++ or Java impose data-racefreedom as a requirement.We propose compiler techniques that automatically delineate extended data-race-free regions (xDRF), namely regions of code which provide the same guarantees as the synchronization-free regions (in the context of DRF codes). xDRF regions stretch across synchronization boundaries, function calls and loop back-edges and preserve the data-racefree semantics, thus increasing the optimization opportunities exposed to the compiler and to the underlying architecture. Our compiler techniques precisely analyze the threads' memory accessing behavior and data sharing in shared-memory, general-purpose parallel applications and can therefore infer the limits of xDRF code regions.We evaluate the potential of our technique by employing the xDRF region classification in a state-of-the-art, dualmode cache coherence protocol. Larger xDRF regions reduce the coherence bookkeeping and enable optimizations for performance (6.8%) and energy efficiency (11.7%) compared to a standard directory-based coherence protocol.

show abstract

“…Sets of concurrent traces containing permutations of events from a given concurrent trace are useful for predictive analysis (e.g., [24,34,35,41]) and synchronization synthesis (e.g., [8,9]) of sharedmemory concurrent programs. Most approaches using such trace sets are restricted to specific aspects of reasoning about concurrent programs such as data race detection [24,34], detection of safety violations [35,41] and fixing assertion failures [8,9].…”

Section: Introductionmentioning

confidence: 99%

“…Most approaches using such trace sets are restricted to specific aspects of reasoning about concurrent programs such as data race detection [24,34], detection of safety violations [35,41] and fixing assertion failures [8,9]. Moreover, the representations of trace sets and exploration strategies used in some of these approaches [8,9,35]) underapproximate the target trace sets.…”

Section: Introductionmentioning

confidence: 99%