Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Learn how to effectively partition concurrent code and manage locking strategies in this conference talk from Kernel Recipes 2025. Explore the fundamental challenge of deciding what data should be protected by which locks and determining the optimal number of locks for your system. Understand why too few locks create performance bottlenecks through contention while too many locks introduce overhead and deadlock risks. Discover approaches for breaking down the "big ball of mud" that characterizes many sequential programs into separate communicating tasks. Gain insights into the broader problem of partitioning processing requirements for concurrent systems and learn practical strategies to make software decomposition more manageable. The presentation addresses this critical aspect of concurrent programming design that affects both performance and maintainability of kernel-level code.
