How does the Linux kernel scheduler decide which process runs next?

01 Kas 21:00

Understanding the decision process behind the Linux kernel's scheduler has always fascinated me. Could someone explain the main criteria the scheduler evaluates when selecting the next runnable task? I'm particularly interested in how priority, load balancing, and real‑time constraints interact, and what role cgroups or nice values play in this selection. Any diagrams or references to the latest scheduling classes would be appreciated. How do you approach studying this subsystem?