Formal specification of MPI 2.0: case study in specifying a practical concurrent programming API

We describe the rst formal speci cation of a non-trivial subset of MPI, the dominant communication API in high performance computing. Engineering a formal speci cation for a non-trivial concurrency API requires the right combination of rigor, executability, and traceability, while also serving as a smooth elaboration of a pre-existing informal speci - cation. It also requires the modularization of reusable speci cation components to keep the length of the speci cation in check. Long-lived APIs such as MPI are not usually `textbook minimalistic' because they support a diverse array of applications, a diverse community of users, and have ef cient implementations over decades of computing hardware. We choose the TLA+ notation to write our speci cations, and describe how we organized the speci - cation of 150 of the 300 MPI 2.0 functions. We detail a handful of these functions in this paper, and assess our speci cation with respect to the aforesaid requirements. We close with a description of possible approaches that may help render the act of writing, understanding, and validating speci cations much more productive.