CVE-2023-0190

What it actually does

NULL pointer dereference in the kernel-mode layer of the NVIDIA GPU Display Driver on Linux that results in a denial of service when triggered.

Why it matters to the business

Unexpected system crashes or loss of GPU acceleration can interrupt workloads on Linux systems using affected NVIDIA hardware, leading to downtime in graphics, compute, or AI environments; low exploit likelihood reduces immediate risk but still requires attention for stability.

How an attack unfolds

An attacker with local low-privileged access on a Linux system crafts input that reaches the vulnerable kernel-mode path in the NVIDIA driver, triggering the NULL dereference and causing a kernel-level crash or freeze; no named threat actors or ransomware campaigns have been associated with this issue and public exploit maturity remains none.

How to fix it

Update the NVIDIA GPU Display Driver to the latest available version that resolves the NULL pointer issue; apply vendor-supplied patches across all affected GeForce, NVS, Quadro, RTX, and Tesla products as soon as feasible.

How to confirm the fix worked

Verify the installed driver version against NVIDIA release notes, reproduce controlled local calls that previously triggered the dereference, and confirm absence of kernel oops or panic messages referencing the driver.

Typical attack path

1. Obtain local user access on an affected Linux host
2. Invoke graphics or compute APIs that exercise the kernel-mode display driver
3. Supply malformed or unexpected parameters that reach the NULL pointer path
4. Trigger the dereference to induce denial of service

How to detect it

Review kernel logs and dmesg output for NVIDIA driver entries containing NULL pointer dereference or oops traces; correlate with process activity that interacts with /dev/nvidia* device nodes.

What your security team should do

• Schedule driver updates during the next maintenance window given the 1273-day publication age and stable low EPSS
• Monitor kernel logs for NVIDIA-related NULL dereference crashes as an interim detection measure
• Restrict untrusted local accounts from accessing GPU devices until patched