CVE-2022-34665

What it actually does

A local user with basic capabilities can trigger a null-pointer dereference in the kernel mode layer of the NVIDIA GPU Display Driver, resulting in denial of service.

Why it matters to the business

Affected systems may experience sudden crashes or loss of GPU functionality from local attacks, interrupting workloads and requiring reboots or restarts with no remote vector involved.

How an attack unfolds

An attacker with local access and standard user privileges crafts input that reaches the vulnerable kernel-mode code path, dereferences a null pointer, and forces the driver to fail, causing system instability or service denial.

How to fix it

Update the NVIDIA GPU Display Driver to the vendor-supplied fixed release for the affected Windows and Linux versions. No apache2 remediation applies to this driver vulnerability.

How to confirm the fix worked

Confirm the installed NVIDIA driver version is at or above the fixed release listed in the vendor advisory; retest local privilege scenarios that previously triggered the dereference.

Typical attack path

1. Obtain local access as a non-privileged user on a Windows or Linux system with the vulnerable NVIDIA driver
2. Invoke GPU-related APIs or operations that reach the kernel-mode display driver
3. Supply malformed parameters that trigger the null-pointer dereference
4. Observe resulting kernel panic, driver crash, or denial-of-service condition

How to detect it

Review system and kernel logs for NVIDIA driver crashes or null-pointer exceptions around the time of reported instability; no network or remote indicators exist.

What your security team should do

• Prioritize patching NVIDIA display drivers on all Windows and Linux endpoints within the next maintenance window
• Audit local user access to GPU resources and restrict unnecessary privileges
• Monitor for driver-related kernel crashes in system logs as an interim detection measure