Eraser

Abstract

Multi-threaded programming is difficult and error prone. It is easy to make a mistake in synchronization that produces a data race, yet it can be extremely hard to locate this mistake during debugging. This paper describes a new tool, called Eraser, for dynamically detecting data races in lock-based multi-threaded programs. Eraser uses binary rewriting techniques to monitor every shared memory reference and verify that consistent locking behavior is observed. We present several case studies, including undergraduate coursework and a multi-threaded Web search engine, that demonstrate the effectiveness of this approach.

Spectrum-Based Fault Localization for Diagnosing Concurrency Faults

“…Static analysis techniques addressing this issue include those based on type systems [6], model checking [16], and general program analysis [17]. There were also dynamic analysis techniques proposed like RecPlay [23] and Eraser [24] However, these techniques were subject to a significant number of false positives. In our approach, we cope with this issue by exploiting a probability score to rank the components instead of providing a binary decision.…”

Static Object Race Detection

“…if root ∈ minBi j [7] if reentrant (i j) [8] St += (j.n, L(i.m j.n)) [9] S t += (j.n, L(i.m j.n)) [10] else [11] St += (j.n, L(i.m j.n)) [12] if j.n not visited then W += j.n…”

“…Their object race detection is dynamic. In fact, a primary goal is to optimize dynamic lockset-based race detection [12]; the higher-level concept of object race entails fewer dynamic checks and therefore lower overhead. In later work, von Praun and Gross introduce the concept of the Object Use Graph (OUG) [16] which allows reasoning about the temporal relation of object accesses, and further reduces the amount of dynamic checks in the lockset-based detector.…”

Multicore Scheduling for Lightweight Communicating Processes

“…This stems from a desire make concurrent programming easier, and to avoid common errors associated with locks and shared-memory [40]. For example message-passing frameworks such as PVM [19] and MPI [21] provide primitives similar to those presented in this paper, but do so for a course-grain network environment.…”