Daniel Riley, a PhD student under Dr. Grigory Fedyukovich, has been awarded a scholarship to attend the Programming Languages Mentoring Workshop (PLMW) at PLDI 2025 in Seoul, South Korea. PLDI, the ACM SIGPLAN Conference on Programming Language Design and Implementation, is a highly regarded event that brings together leading experts and emerging talent in programming language research.
Riley’s participation in PLDI 2025 will be further highlighted by the presentation of the paper, “Exact Loop Bound Analysis”. This research tackles the largely unexplored area of exact loop bound analysis, a technique that offers new applications for precisely determining the number of iterations a program loop will execute. Unlike traditional methods that focus on upper or lower bounds, this work introduces a novel approach to find an exact bound by iteratively considering under-approximations of a program, synthesizing a function that, when applied to initial program variable values, yields an exact bound. The algorithm synthesizes a function to capture all possible execution cases for a loop and provides a way to compute the exact number of iteration the loop will perform. The approach utilizes data analysis and abductive inference to find each piece of the exact bound, and each piece is checked by a verification engine to ensure its correctness. This innovative technique, implemented in a tool called ELBA, has shown effectiveness on challenging single-loop benchmarks and is poised to provide a more precise understanding of program execution costs, a critical factor for optimizing performance on limited hardware and guaranteeing program execution. The algorithms code can be used in an artifact made available at [https://zenodo.org/records/15161619](https://zenodo.org/records/15161619).
The scholarship will enable Daniel to engage with top researchers and further his work in this area of study in computer science.