NVIDIA module missing after a kernel update · Linux Fix Lab

Symptoms & scope

The desktop falls back to a basic display or the display manager fails after an update.
modprobe reports that the NVIDIA module cannot be found for the running kernel; nvidia-smi may also fail.

Relevant environment

Ubuntu and related distributions using packaged NVIDIA kernel modules or DKMS; Fedora uses a different akmods packaging path. Applies only when evidence actually names a missing module.

Recognizable messages (synthetic examples)

modprobe: FATAL: Module nvidia not found in directory /lib/modules/7.0.0-30-generic

Check the running kernel and matching module package before diagnosing signatures, hardware or Vulkan.

Possible causes

These are possible explanations, not a confirmed diagnosis. Several independent faults can coexist.

A precompiled NVIDIA module package may be missing for the kernel ABI that actually booted, even though an older version remains installed.
A DKMS build may not have completed for that kernel because headers or build dependencies were missing. This differs from Secure Boot rejecting an existing unsigned module.

Diagnose safely

Run one command at a time in the relevant session. Read the explanation first. Uppercase placeholders need your own values; tools and privileges vary by distribution. These commands are displayed here and never executed by the website.

Check 1

Read the currently running kernel release without installing packages.

uname -r

Interpret the result: A packaged NVIDIA module must match this kernel, not just the newest installed kernel.

Check 2

Ask kmod for the module file resolved against the current kernel; no module is loaded by this check.

modinfo -n nvidia

Interpret the result: A missing-path failure supports a module availability problem. A path by itself does not establish a successful load or correct userspace version.

Check 3

Use only where DKMS is installed; on Ubuntu a precompiled module can exist without a DKMS entry.

dkms status

Interpret the result: Built/installed records refer to particular kernel versions. No DKMS entry does not prove the NVIDIA stack is broken.

Evidence-guided next steps

Compare installed kernel ABI and matching module package

On Ubuntu, inspect the installed linux-modules-nvidia package for the active driver branch and exact kernel ABI using the distribution package manager. Check whether the release uses precompiled modules or a DKMS workflow before proposing any package action. Prefer Canonical's supported driver installer and an exact branch match instead of downloading random .run installers.

Precautions: Keep the previously working boot entry available; do not purge all graphics packages based on a failed nvidia-smi alone.

Recovery / rollback: Boot the previously working kernel if needed; undo only a precisely identified package change.

Did this solution help you?

Share this solution #

Check Secure Boot separately from missing modules

If modinfo finds a module but loading still fails, examine journal messages and its signer before treating it as absent. Ubuntu Secure Boot validates third-party module signatures; an existing module rejected by trust policy calls for a signing or MOK investigation, not reinstallation of arbitrary libraries. A module that does not exist for the booted kernel cannot be made present merely by disabling Secure Boot.

Precautions: Never make disabling Secure Boot the default repair; it changes a security boundary.

Recovery / rollback: Keep the original Secure Boot configuration and use the distribution's documented MOK process only if signed modules are the established issue.