Causality-based Verification
Andrey Kupriyanov
Program verification is one of the central research topics in computer science since its inception – we can consider the field to be initiated as early as in 1949, with Alan Turing’s pioneering paper “Checking a Large Routine.” Yet, we are still far from the dream of automatically proving every computer program correct. Two aspects make this problem particularly challenging: concurrent program execution on parallel processors, and large, or even infinite, state spaces of data-manipulating programs. Nowadays, with concurrency entering everywhere, from smartphones to aircrafts, proving the correctness of infinite-state concurrent programs becomes increasingly more important: we do want to be sure that the program that controls the airplane we are flying in is correct.
In this thesis we propose a new approach to the verification of infinite-state concurrent programs. We call it causality-based, because it captures in an automatic proof system the “cause-effect” reasoning principles, which are often used informally in manual correctness proofs. While traditionally automatic methods are based on the state space exploration, our method is based on a new concurrency model, called concurrent traces, which are the abstractions of the history of a concurrent program to some key events and the relationships between them. Causality-based proof rules relate concurrent traces with each other, by formally tracking what are the necessary consequences (the “effects”) from a particular analysis situation (the “cause”). The full correctness proof is then a composition of such primitive proof steps.
We study the syntactic and language-based properties of concurrent traces, and characterize the complexity of such operations as emptiness checking and language inclusion. Regarding the program correctness, we develop proof systems for the broad classes of safety and liveness properties, and provide algorithms for the automatic construction of correctness proofs. We demonstrate that for practically relevant classes of programs, such as multi-threaded programs with binary semaphores, the constructed proofs are of polynomial size, and can be also checked in polynomial time. The methods of the thesis have been implemented in Arctor, the first scalable termination prover for concurrent programs, which is able to handle programs with hundreds of non-trivial threads.
PhD Thesis.