| Lesson 12 | Linux Kernel Configuration and Compilation: Module Summary |
| Objective | Review module management, the source-build workflow, and the checks needed before trusting a newly booted kernel. |
This module connected everyday kernel administration with the more specialized task of building a kernel from source. The first decision is whether a custom build is needed at all. For routine RHEL administration, supported kernel and driver packages are the normal starting point. The upstream build exercises belong in a prepared lab with a recovery path.
modprobe to resolve dependencies, verify results, and remove modules only when their users allow it.modprobe.d from deliberate startup requests in modules-load.d..config, boot dependencies, signing inputs, and a distinct local-version suffix.Available versus loaded: a module can exist on disk without being loaded. A built-in driver can work without appearing in lsmod. Use modinfo and configuration information alongside the loaded-module list.
Running versus target: uname -r describes the kernel executing now. Before rebooting, use the recorded target release to inspect the new module directory, image, and initramfs.
Built versus verified: a successful build produces artifacts. A working installation also needs a usable boot entry, trusted code under the system's policy, access to the root filesystem, and successful workload checks.
For a lab build, retain the source release or revision, any patches, the configuration, relevant tool versions, target release, installation method, and test results. After selecting the target in the boot menu, a useful initial check is:
uname -r
lsmod
findmnt /
sudo journalctl -k -b --no-pager
Then test the feature that motivated the build and the system services that depend on the kernel. These commands provide evidence, but no single command certifies that every driver and application works. Keep the previous kernel until the new system has passed the checks appropriate to its use.
.config that select features and dependencies.lsmod?You can now approach kernel changes as a controlled sequence: establish the need, configure deliberately, build reproducibly, install matching artifacts, and verify the boot and workload. Continue to the next module when you are comfortable explaining that sequence.