NVIDIA GPU driver for Linux: memory leak in error paths exhausts kernel memory

CVSS 5.5CVE-2026-47566NVIDIA / GPU stackcurated

Impact

An unprivileged local user can repeatedly drive the driver down error paths that fail to free memory, until kernel memory is exhausted. This is a slow denial of service that ends in the whole node, not just the GPU, becoming unusable - and it looks like a mysterious OOM rather than an attack, so it can burn operator time before anyone suspects a tenant. The guest driver and vGPU manager are both affected. Availability only.

Who can reach it

Any local unprivileged user with access to the NVIDIA driver, including a tenant GPU pod or a guest VM.

What to do

Update the Linux GPU display driver, guest driver and vGPU manager to the fixed branch in NVIDIA bulletin 2026/5861. Drain the node and reboot. Until patched, node-level memory alerting is what gives you warning; there is no documented mitigation.

References

https://nvd.nist.gov/vuln/detail/CVE-2026-47566 https://github.com/NVIDIA/product-security/tree/main/2026/5861 https://www.cve.org/CVERecord?id=CVE-2026-47566

Related entries

All NVIDIA / GPU stack entries

This entry is curated: imported from vendor advisories with machine assistance, not yet individually verified. Confirm against your vendor's advisory before acting, and report anything wrong.