New eBPF Tool Slashes Software Dependency Validation Overhead

Computer scientists at Waseda University have developed a new tool, mkcheck2, designed to drastically reduce the time and compute costs associated with software build dependency verification. According to reporting from The Register, researchers Yuta Saito, Kazunori Sakamoto, and Hironori Washizaki found that dependency specification management is a persistent pain point, contributing to more than half of all build errors in large-scale projects.

To address this, the team moved away from existing ptrace-based tools, which impose significant overhead due to required process suspensions and context switches. To improve efficiency, mkcheck2 implements system call tracing based on the extended Berkeley Packet Filter (eBPF). By running sandboxed programs directly within the kernel space, the tool provides non-invasive tracing with minimal impact on build performance.

In a paper published in the Proceedings of the 2026 IEEE/ACM 48th International Conference on Software Engineering, the authors report that mkcheck2 reduces dependency error detection overhead by up to 99.7% compared to ptrace-based methods. When tested against a 300-project Make corpus, the tool's incremental analysis technique reduced the mean analysis time per commit from 1,267.49 seconds to 23.56 seconds—a speed increase of approximately 54x. This efficiency gain is intended to make continuous dependency verification practical for real-world development environments.

The researchers noted certain limitations: the eBPF-tracing system is specific to Linux. Additionally, challenges remain regarding distributed build systems, network dependencies, visibility into dynamically loaded libraries, and tracking access to memory-mapped regions.