Abstract:With the development of network technology and the increasing popularity of distributed systems, the society pays more and more attention to the reliability of distributed systems. Distributed systems establish services and communications among a large number of terminals, and message passing is a common communication method in distributed systems. In practice, asynchronous mode has better performance than synchronous mode, but asynchronous completion makes the task of specifying correct behaviors more difficult. Verification of asynchronous distributed systems is challenging due to unpredictable interleaving and possible network failures. Aiming to verify the safety properties of asynchronous programs, we propose a simple procedure with the assumption of mergeable parallelism in this paper. Then we characterize inference rules to describe the sequence combination that satisfies the conditions of the receive operations. The program’s execution can be reduced to executions with sets of fixed order. A proof is provided to show its soundness. The correctness of the mergeable message passing programs could be verified by a state-of-the-art verification framework. Experimental results show that our method is efficient and applicable in various message passing cases.

Symbolic Verification of MPI Programs with Non-deterministic Synchronizations.

MPI-SV: a symbolic verifier for MPI programs

Symbolic Verification of Message Passing Interface Programs

Symbolic Verification of Message Signatures in MPI

Synchronization Error Detection of MPI Programs by Symbolic Execution

MPISE: Symbolic Execution of MPI Programs.

Symbolic Execution of MPI Programs with One-Sided Communications

Combining Model Checking and Testing with an Application to Reliability Prediction and Distribution

Towards deductive verification of MPI programs against session types

Deterministic Message Passing for Distributed Parallel Computing

Verifying the Correctness of Distributed Systems Via Mergeable Parallelism

Evaluation of Model Checkers by Verifying Message Passing Programs

Verifying the safety properties of distributed systems via mergeable parallelism

dMPI: Facilitating Debugging of MPI Programs via Deterministic Message Passing.

Verification of MPI programs

Verification of Message-Passing Uninterpreted Programs

Trace-Based Temporal Verification for Message-Passing Programs

A symbolic approach for detecting conflicts in Verilog's non blocking assignments

Towards a Framework for Scalable Model Checking of Concurrent C Programs

Verifying Parallel Low-Level Programs for Multi-core Processor

Methods to Model-Check Parallel Systems Software