Herding Cats Modelling, Simulation, Testing, and Data Mining for Weak Memory

• Published in: ACM transactions on programming languages and systems Vol. 36; no. 2; pp. 1 - 74
• Main Authors: Alglave, Jade, Maranget, Luc, Tautschnig, Michael
• Format: Journal Article
• Language: English
• Published: New York, NY, USA ACM 01.07.2014; Association for Computing Machinery
• Subjects: Applied sciences; Computer Science; Computer science; control theory; systems; Computer simulation; Computer systems and distributed systems. User interface; Computer systems organization; Computer systems performance. Reliability; Consistency; Data processing. List processing. Character string processing; Distributed, Parallel, and Cluster Computing; Equivalence; Exact sciences and technology; Hardware; Integrated circuits; Mathematical models; Memory organisation. Data processing; Modelling; Programming Languages; Semiconductor memory; Simulation; Software; Specifications; Stores; Concurrency; weak memory models; software verification; Data analysis; Computer simulation; Theory; Finite state machine; Verification; Linux system; Distributed system; Data mining; C language; Modeling; Axiomatic; Program verification; Storage management; Sequential consistency; Model specification; Model checking; UNIX system; Experimentation; Anomaly; Simulator; User behavior; C0 [Hardware/Software Interfaces] General Terms: Theory; Verification Additional Key Words and Phrases: Concurrency; Categories and Subject Descriptors: B32 [Shared Memory]
• ISSN: 0164-0925; 1558-4593
• DOI: 10.1145/2627752

We propose an axiomatic generic framework for modelling weak memory. We show how to instantiate this framework for Sequential Consistency (SC), Total Store Order (TSO), C++ restricted to release-acquire atomics, and Power. For Power, we compare our model to a preceding operational model in which we found a flaw. To do so, we define an operational model that we show equivalent to our axiomatic model. We also propose a model for ARM. Our testing on this architecture revealed a behaviour later acknowledged as a bug by ARM, and more recently, 31 additional anomalies. We offer a new simulation tool, called herd, which allows the user to specify the model of his choice in a concise way. Given a specification of a model, the tool becomes a simulator for that model. The tool relies on an axiomatic description; this choice allows us to outperform all previous simulation tools. Additionally, we confirm that verification time is vastly improved, in the case of bounded model checking. Finally, we put our models in perspective, in the light of empirical data obtained by analysing the C and C++ code of a Debian Linux distribution. We present our new analysis tool, called mole, which explores a piece of code to find the weak memory idioms that it uses.