📌 Prerequisites

Before this lecture you should have completed Lectures 1–7 in this free Linux kernel development course, covering basic Linux command-line usage, the role of the kernel, and how user space and kernel space are separated. Familiarity with the ls, cd, head, and grep commands is assumed.

💡 Real-World Analogy — The Architect’s Blueprint Set

Think of a large building under construction. The contractor has a complete set of blueprints — electrical plans, plumbing diagrams, structural drawings, interior layouts. Each type of plan is in its own labelled folder. The Linux kernel source tree works the same way. The finished building is the running kernel; the blueprint folder set is the source tree. You do not need to read every blueprint to walk inside the building, but a kernel developer absolutely must know which folder holds which plan.

💡 Why Is the Compiled Kernel Much Smaller Than the Source?

The source tree contains code for every CPU architecture (x86, ARM, RISC-V, MIPS, and more), every supported device driver, all filesystems, and every optional feature. When you build a kernel you use a configuration file (.config) that selects only the features your specific system needs. Everything not selected is not compiled. The source is the complete universe of possibilities; the compiled image is just the chosen subset for one system.

🎉 What is the NAME field?

Each kernel release gets a fun, often absurd nickname picked personally by Linus Torvalds. Kernel 5.4 was called “Kleptomaniac Octopus”. Kernel 6.1 was “Tantric Tasmanian Tiger”. Kernel 6.6 was “Pinguino Sincero”. These names have no technical meaning but they do make version discussions a bit more memorable.

🚀 rust/ — New in kernel 6.1, stable in kernel 7.0

This directory did not exist in kernel 5.x. It contains the Rust programming language infrastructure for the kernel — the abstractions, bindings, and helper code that allow new kernel components to be written in Rust. As of Linux 7.0 (2025), Rust is no longer experimental. Rust-written drivers covering PCI enumeration, interrupt handling, DMA, and platform devices are fully production-grade.

⚡ io_uring/ — Promoted to top-level in Linux 5.1+

The io_uring subsystem is Linux’s modern high-performance asynchronous I/O interface. It uses shared ring buffers between user space and the kernel to allow applications to submit and receive I/O results without making a system call per operation. It has grown so central to Linux performance that it now occupies its own top-level directory.

📌 Important for Students of This Free Linux Kernel Course

Learning C for kernel development is still absolutely essential. The vast majority of the kernel is C and will remain so for many years. Rust is an addition, not a replacement. The first thing this free embedded systems course covers is C-based kernel and driver writing. Rust becomes relevant once you are comfortable with the kernel model itself.

🔎 Pro Tip — Use Elixir Cross-Reference Online

Instead of running grep locally across a 1.5 GB source tree, use elixir.bootlin.com. It is a free web-based cross-reference tool for the Linux kernel. You can search any function name, macro, or struct and instantly see every file where it is defined or used, across multiple kernel versions. It is invaluable for kernel learning and development.

Q1. What is the Linux kernel source tree?

The Linux kernel source tree is the complete, hierarchically organised collection of source code that compiles into the Linux kernel. It contains code for every supported CPU architecture, device driver, filesystem, network protocol, and kernel subsystem. It is officially hosted at kernel.org and the entire history is managed in a Git repository.

Q2. Do I need to download the full git history to start kernel development?

No. For learning or building your first kernel, download the tarball from kernel.org — it is a compact, clean starting point. A git clone --depth=1 is the next step up. The full git clone (all history) is only needed when you are studying the evolution of specific code or preparing to upstream a patch and need access to git log, git blame, and the full commit history.

Q3. Why is there a NAME field in the Makefile like “Kleptomaniac Octopus”?

Linus Torvalds gives each kernel release a playful nickname. It has no technical meaning but it makes it easy to refer to a release informally. The name is set in the top-level Makefile under the NAME variable and is visible when you run head Makefile from the source root.

Q4. Should I learn Rust to contribute to the Linux kernel?

C is still the dominant kernel language and will remain so for years. You must learn C-based kernel programming to work effectively anywhere in the existing kernel tree. Rust knowledge becomes a differentiator for writing new drivers where memory safety is a priority. For students following this free Linux kernel development course, focus on C first — Rust will be introduced once you understand the kernel programming model.

Q5. What does EXTRAVERSION in the Makefile tell me?

EXTRAVERSION is a suffix appended after SUBLEVEL. For a release candidate it will be something like -rc4. For a final stable release it is empty. Linux distributions also add their own suffix here — Ubuntu kernels might show -generic and Raspberry Pi kernels might show -v8+. This is purely informational for version identification.

Q6. Is this Linux kernel development course really free?

Yes. EmbeddedPathashala provides this entire Linux kernel programming series, along with courses on Linux device drivers and embedded systems, completely free. The goal is to make high-quality technical education available to every student regardless of background or budget. All lectures, diagrams, and code examples are provided at no charge.

Q7. How do I find who maintains a particular driver or subsystem?

Use the script at scripts/get_maintainer.pl from the kernel source root. Pass it the path to any file and it reads the MAINTAINERS file to tell you exactly which person or team is responsible for that code, which mailing list to use for patches, and the status of that component. For example: ./scripts/get_maintainer.pl drivers/net/ethernet/intel/e1000e/e1000.h

Q8. Can I browse the kernel source without downloading it?

Yes, two excellent options: elixir.bootlin.com provides cross-referenced, searchable access to the kernel source across many versions. github.com/torvalds/linux is a read-only mirror of the official kernel repository and supports the familiar GitHub file browsing interface. Neither replaces having a local copy when you are ready to build and test changes.

Q9. What is io_uring and why does it have its own top-level directory now?

io_uring is Linux’s modern asynchronous I/O interface. It uses a pair of ring buffers shared between the kernel and user space — a submission queue where the application writes I/O requests, and a completion queue where the kernel writes results. Because both are memory-mapped, applications can submit many requests and collect results without a system call per operation, dramatically reducing I/O overhead for high-throughput workloads. It has grown so important that it now has its own top-level directory, separate from the older async I/O code.

Q10. What is the best way to search inside the kernel source tree?

For quick searches locally, use grep -rn "search_term" --include="*.c" from the source root. For symbol cross-referencing, use elixir.bootlin.com. For understanding a subsystem, start with Documentation/ — the relevant subdirectory almost always has a README or rst documentation file that explains the subsystem design. For example, Documentation/filesystems/ covers the VFS and individual filesystem internals.