Context switching is necessary for a high-throughput and responsive system where all processes make progress despite limited execution resources. But, as we discussed in the previous article, it also has performance costs which cascade through various indirect mechanisms, such as cache and TLB eviction.

When building performance-critical systems or debugging performance issues due to context switching, it becomes important to understand the internal implementation details to be able to reason through the performance issues and possibly mitigate them. Not only that, it leads you to learn many low-level details about the hardware architecture, and makes you realize why the kernel is so special.

At first glance, context switching seems straightforward—save the current process's registers, switch page tables and stacks, and restore the new process's registers.

However, the reality is much more complex, involving multiple data structures, hardware state management, and memory organization. To fully grasp context switching, we need to understand few key foundational concepts about the Linux kernel and X86-64 architecture.