Compiler analysis for OpenMP tasks correctness

Abstract

OpenMP has become the most frequently used programming model for shared memory parallel systems by virtue of its simplicity and scalability. Notwithstanding its easiness, the constant evolution of OpenMP brings forth difficulties when it comes to keep up with its capabilities. Compilers are key tools to anticipate bugs that may appear at runtime. Despite this, most compilers do not diagnose common mistakes that cause execution errors. Furthermore, many compilers do not yet implement the latest extensions of the OpenMP asynchronous model, and there are no checking systems to properly identify errors in that model.In this paper, we focus on the mistakes derived from the usage of OpenMP tasks. We define a data-flow algorithm that computes the live tasks at every point of the program. Based on that, we present different scenarios and the analyses that allow detecting possible runtime failures, loss of performance or non-deterministic results. We have implemented all the analyses in the Mercurium source-to-source compiler, which, based on the results, suggests changes to the user in order to avoid runtime problems. We test our implementation with over 70 students and 6 benchmarks, as well as with the results of the Oracle Solaris Studio 12.3 compiler, which warns about correctness issues related to the scope of variables in OpenMP.

Towards a Qualifiable OpenMP Framework for Embedded Systems

“…OpenMP tools do not usually take into account safety requirements due to its HPC orientation. However, several works consider safety a major issue even in OpenMP, and tackle the problem from a specification perspective, proposing modifications in order to provide a more robust and safe language, e.g., enhancing programmability [11], [12], detecting data races and other data environment errors [13], [14], and improving resiliency [15]. More generally, there are works that consider the analysis of the functional safety [16] and the time predictability [5], [6] of the whole OpenMP specification, both concluding that OpenMP can be functionally safe and time predictable including only minor changes.…”

Improving the performance of fragments of computer code with high time costs based on parallel programming technologies

“…Директива reduction обозначает переменную, с которой в цикле производится reduction-операция (например, суммирование). При выходе из цикла, данная операция производится над копиями переменной во всех потоках, и результат присваивается оригинальной переменной [13].…”

Safe Parallelism: Compiler Analysis Techniques for Ada and OpenMP

“…Among the latter, Ma et al created a tool for detecting race conditions in OpenMP programs with a fixed number of threads [22], and Basupalli et al developed a robust technique for detecting race conditions in OpenMP programs using affine constructs [23]. Finally, Royuela et al developed a series of algorithms focused on the OpenMP tasking model to find incoherences in data-sharing and dependence clauses, as well as race conditions [24].…”