CVE-2024-0091

What it actually does

The vulnerability permits a local user to trigger an untrusted pointer dereference through a driver API call in the NVIDIA GPU Display Driver, resulting in memory corruption that can cause denial of service, information disclosure, or data tampering.

Why it matters to the business

Systems running the affected driver on Windows or Linux may experience crashes, unauthorized data exposure, or unauthorized modifications, leading to downtime, compliance risks, and potential loss of sensitive information; the issue affects a broad range of GPU-equipped endpoints with low but stable exploit probability.

How an attack unfolds

An attacker with local access on a vulnerable system crafts a call to the exposed driver API that supplies an untrusted pointer; successful dereference corrupts memory structures, enabling the attacker to crash the driver, read arbitrary memory, or alter data.

How to fix it

Apply the vendor-supplied fixes immediately: update gitlab-org/gitlab to commit c8f3448a2f0cfb3c812fc4619e0f16bc3feb133b and electron-userland/electron-builder to commit 3f3419a483bcc43018ddb0c9fdeb38f6abdba84a; where NVIDIA driver packages are present, obtain and install the corresponding patched driver release from the vendor.

How to confirm the fix worked

Verify that the listed commit hashes or newer driver builds are deployed, recompile or reinstall affected packages, and confirm through driver version checks or API testing that the pointer dereference path no longer accepts untrusted input.

Typical attack path

1. Gain local user access on a target system with the vulnerable NVIDIA driver installed
2. Load or invoke the affected driver API with a maliciously chosen pointer value
3. Trigger the pointer dereference to induce memory corruption
4. Escalate the corruption into denial of service, data leakage, or tampering

How to detect it

Review system and driver logs for unexpected crashes or API failures around the time of suspected activity; examine process execution traces for repeated calls to the affected driver entry points.

What your security team should do

• Prioritize patching gitlab and electron-builder instances to the specified commits within the next maintenance window
• Inventory all Windows and Linux systems using NVIDIA GPU drivers and schedule driver updates
• Monitor EPSS and exploit maturity indicators for any future increase in activity
• Restrict local user access to driver APIs where feasible through least-privilege controls