---
id: "CVE-2026-99224"
kind: "vulnerability"
aliases: ["CVE-2026-99224"]
title: "Linux kernel: key describe can read freed module memory and crash"
severity: "high"
cvss: 7.0
epss: null
exploited: false
status: "open"
category: "infrastructure"
products: []
ecosystems: []
tags: ["af_rxrpc", "infrastructure", "kernel", "keyctl", "keyring", "linux"]
not_affected_if: ["af_rxrpc-not-built-as-module", "af_rxrpc-not-loaded", "single-user-host"]
published_at: "2026-11-02T16:17:46+00:00"
first_seen_at: "2026-11-02T16:24:05+00:00"
updated_at: "2026-11-02T16:29:31+00:00"
sources: [{"source": "nvd", "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-99224", "first_seen_at": "2026-11-02T16:24:05+00:00"}, {"source": "analyst", "url": null, "first_seen_at": "2026-11-02T16:29:31+00:00"}]
---

# Linux kernel: key describe can read freed module memory and crash

A race in the Linux kernel keyring code lets a local task read the type name of a key after its module has been unloaded, causing a kernel page fault. The sources describe it as a kernel oops in the key description path, triggered when the AF_RXRPC module is unloaded during a keyctl describe call.

_Summary, check and fix written by a model from the sources; labelled as such._

## How to check

- Check whether af_rxrpc is a loadable module: `grep CONFIG_AF_RXRPC /boot/config-$(uname -r)` (=m is the affected build) and `lsmod | grep rxrpc`
- Not affected by this path if AF_RXRPC is built in (=y) or absent.
- Check your distro's kernel changelog for the commit above or the CVE id; `uname -r` alone cannot settle it until fixed versions are published.

## Fix

Upstream fix is commit fced7238b7e80cf8aa688d5af976e4aa6e33c329 ('keys: Protect the type name while describing a key', security/keys/keyctl.c, +2 lines). Fixed stable release versions: unknown. Check your distro kernel changelog for this commit or CVE-2026-99224.

## Timeline

- 2026-11-02T16:24:05+00:00 nvd https://nvd.nist.gov/vuln/detail/CVE-2026-99224
- 2026-11-02T16:29:31+00:00 analyst — Read the upstream fix commit fced7238b7 (Cen Zhang, reviewed by Jarkko Sakkinen, Cc stable, 2026-10-09): holds key->sem for read across kasprintf() in keyctl_describe_key(). Fixed release numbers and stable backport commits are NOT in the sources I found; NVD page fetch returned empty. Nothing marked verified: the commit is a mirror (upd.dev), not kernel.org.

## Documents

- [describes] NVD CVE-2026-99224 https://nvd.nist.gov/vuln/detail/CVE-2026-99224
- [discussion] Re: [PATCH] keys: Protect the type name while describing a key (LKML) https://lkml.iu.edu/2610.1/02319.html
- [fix] keys: Protect the type name while describing a key (fced7238b7) https://upd.dev/torvalds/linux/commit/fced7238b7e80cf8aa688d5af976e4aa6e33c329

## From the sources

In the Linux kernel, the following vulnerability has been resolved:

keys: Protect the type name while describing a key

keyctl_describe_key() must keep the key type's name alive until
kasprintf() has consumed it. The retained key reference keeps the key
and its description alive, but does not pin the module providing its
type. The name is passed to kasprintf() without holding key->sem.

kvasprintf() uses the saved name pointer in two formatting passes, with
a GFP_KERNEL allocation between them. With AF_RXRPC=m, a task that
passes View permission before type retirement can overlap a privileged
module unload in this order:

1. KEYCTL_DESCRIBE looks up the key, saves key->type->name in the
   kasprintf() arguments and completes the first formatting pass.
2. While the formatter is delayed before its second pass,
   af_rxrpc_exit() calls unregister_key_type(), which removes the type
   from the registry and waits for key_gc_keytype().
3. The collector takes key->sem for writing, retypes the retained key
   to key_type_dead and completes retirement. Unregistration and module
   exit return, allowing free_module() to release the type and name.
4. The formatter resumes its second pass and reads the saved name from
   released module storage, which can fault in string().

Hold key->sem for reading across kasprintf(). The collector then cannot
retype the key or complete retirement until both formatting passes have
finished. If retirement wins the lock first, formatting uses the core
key_type_dead name instead. Release the semaphore immediately after
kasprintf(), including on allocation failure, before copying to
userspace.

Oops report as below:

    BUG: unable to handle page fault for address: fffffbfff8054794
    #PF: supervisor read access in kernel mode
    #PF: error_code(0x0000) - not-present page
    RIP: 0010:string+0x23f/0x470
    Call Trace:
     vsnprintf+0x330/0x1110

---

_Free to read at a human pace. An API with keys and an MCP server are in preparation; early access: hello@infrastack.ai_