CVE-2026-99224

A flaw was found in the Linux kernel's key management subsystem. A local user can cause a Denial of Service (DoS) or potentially achieve arbitrary code execution by exploiting a race condition when retrieving key information while the corresponding kernel module is being unloaded. This issue occurs because the kernel accesses key type information without proper locking, leading to a use-after-free condition and a system crash.

Red Hat has evaluated this vulnerability as a Moderate severity impact because triggering the underlying condition requires precise race timing against an administrative kernel module removal, restricting the threat to authenticated local environments. In typical enterprise deployments, kernel modules are loaded at boot or on-demand and rarely unloaded dynamically during routine operations. Furthermore, unprivileged users lack the administrative capabilities necessary to initiate module removal, making deliberate exploitation unlikely under standard operating conditions.

The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.

--- CME Mitigation Guidance (2026-11-04) --- Disclaimer: The following mitigation solutions are provided as a courtesy, pending the application of an official patch. This content was generated utilizing AI tools in conjunction with data from https://cmetaxonomy.org. Customers are advised to independently evaluate and test these temporary mitigations within their own infrastructure.

The highest-value action is to block unprivileged invocation of keyctl syscalls via seccomp or disallow runtime unloading of kernel modules that register key types: upgrade to the corresponding fixed Red Hat builds, which hold key->sem for reading across the kasprintf call and release it before copying to userspace, closing race condition between KEYCTL_DESCRIBE and module unloading. Until patched, this is only a concern on Linux systems where local users have View permission on keys backed by dynamically unloadable kernel modules — non-Linux operating systems are unaffected — so prevent routine dynamic unloading of kernel modules and restrict untrusted local user permissions to create or view keys. As a compensating measure, enable runtime memory error detection using KASAN or HWASan to detect use-after-free memory access, enforce SELinux policy to restrict key security classes and deny unauthorized View permissions, map local users to confined SELinux domains lacking permissions to describe kernel keyrings, and deploy gVisor sandboxing to intercept syscalls and isolate the host kernel key subsystem. These controls limit blast radius and contain exploitation, but they do not fix the underlying locking race use-after-free defect — apply the upstream patch to fully remediate.

Highest-impact CME: CME-601 Kernel-Level Syscall Filtering (seccomp) (high)