GPU VulnDB

Database/Control plane, storage & DevOps

Linux kernel NFSD: revoking pNFS layout state loops forever and hangs the revoker

CVSS 5.5CVE-2026-64519Control plane, storage & DevOpscurated

Impact

When an NFS server revokes state for a filesystem, layout stateids are never marked as revoked, so the revocation loop finds the same layout on every pass and never finishes. The task doing the revocation hangs, and the NFS server stays stuck partway through the operation. Revocation is what HA NFS failover does when it unlocks an export before moving it, so on a dataset or home-directory server for a GPU cluster this can stall a failover. Only servers that hand out pNFS layouts are affected. Plain NFSv3 and v4 exports without layouts never reach this code.

Who can reach it

Local and privileged. The trigger is an administrative state revocation (for example a write to the nfsd unlock interface during failover) on a server with outstanding pNFS layouts. The record describes no path for a remote or unauthenticated user.

What to do

Update to a stable kernel that carries the fix commits referenced below, then reboot the NFS server. Clients see an outage unless you fail over to a patched peer first. Until then, avoid revoking state on exports that have active pNFS layouts. The record does not name fixed kernel versions.

References

Related entries

All Control plane, storage & DevOps 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.