An Efficient Algorithm for On-the-Fly Data Race Detection Using an Epoch-Based Technique

Abstract

Data races represent the most notorious class of concurrency bugs in multithreaded programs. To detect data races precisely and efficiently during the execution of multithreaded programs, the epoch-based FastTracktechnique has been employed. However, FastTrackhas time and space complexities that depend on the maximum parallelism of the program to partially maintain expensive data structures, such as vector clocks. This paper presents an efficient algorithm, callediFT, that uses only the epochs of the access histories. Unlike FastTrack, our algorithm requiresO(1)operations to maintain an access history and locate data races, without any switching between epochs and vector clocks. We implement this algorithm on top of the Pin binary instrumentation framework and compare it with other on-the-fly detection algorithms, including FastTrack, which uses a state-of-the-art happens-before analysis algorithm. Empirical results using the PARSEC benchmark show thatiFT reduces the average runtime and memory overhead to 84% and 37%, respectively, of those of FastTrack.

AR653: Dynamic Detection Tool for Atomicity Races in ARINC 653 Applications

“…However, it is difficult to use these information and to check exist atomicity races in the applications, because detecting atomicity races requires to understand parallel executions of processes and to predict their nondeterministic behaviors. Therefore, in general, a range of automatic detection tools based on sophisticated techniques [10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25][26][27] is employed to locate atomicity races which exist in ARINC 653 applications. A representative tool for detecting atomicity races in the avionics application is CodeSonar.…”

Efficient Filtering Technique for Reducing Time Overhead of Dynamic Data Race Detection in Multithread Programs

“…This paper presents a loop filtering technique that rules out repeatedly execution regions of loops from the monitoring targets to provide a minimum runtime overhead. We evaluated the filtering technique under a dynamic data race detection technique, FastTrack [8] and RaceChaser [9], with multithread programs using a huge amount of loop. The empirical results using multithread programs show that the filtering technique reduces the average runtime overhead to 40% of that of dynamic data race detection.…”

“…Dynamic analysis employ trace based post-mortem methods or on-the-fly. On-the-fly methods are based on three different analysis methods: lockset analysis [10][11], happens-before analysis [8,9,12,13], and hybrid analysis [14][15]. The lockset analysis is simple and can be implemented with low overhead.…”

ConpathView: A Visualization Tool for Debugging Race Conditions in Event Synchronization of ARINC 653 Applications

“…However, if not properly used, this service may cause non-deterministic process executions and cannot guarantee their execution order to shared resources. Since this may cause concurrency errors, such as race conditions [5][6][7] which lead to unpredictable results, it must be considered for debugging. It is highly difficult to identify race conditions caused by event synchronization, because a huge amount of information should be analyzed in combination including the aspect of non-deterministic executions and accesses to the shared memories.…”