Detecting threadsafety violations in hybrid openmp/mpi programs.

Abstract-We propose an approach by integrating static and dynamic program analyses to detect threadsafety violations in hybrid MPI/OpenMP programs. We innovatively transform the thread-safety violation problems to race conditions problems. In our approach, the static analysis identifies a list of MPI calls related to thread-safety violations, then replaces them with our own MPI wrappers, which involve accesses to some specific shared variables. The static analysis avoids instrumenting unrelated code, which significantly reduces runtime overhead. In the dynamic analysis, both happen-before and lockset-based race detection algorithms are used to detect races on these aforementioned shared variables. By detecting races, we can identify thread-safety violations according to their specifications. Our experimental evaluation over real-world applications shows that our approach is both accurate and efficient. designed a practical tool to utilize the on-the-fly dynamic monitoring to detect data races in OpenMP programs. Kang et al [6] presented a tool that focuses on the detection of first data races with no explicit happen-before order in OpenMP programs. The paper II. OVERVIEW This paper proposes a novel approach to detect thread-safety violations using the integrated static and dynamic program analysis. III. STATIC ANALYSIS To determine whether two MPI calls can be concurrently executed at thread level, as shown below, we replace MPI routines with wrappers that can catch runtime information in MPI calls, such as source, tag, communicator, and thread ID information besides executing the original MPI routine