Disclaimer: This content was generated using AI. While we strive for accuracy, please verify critical information with official sources.
Sources: National Vulnerability Database, CISA KEV, Exploit-DB
CVE-2026-93211 Overview
CVE-2026-93211 is a Linux kernel vulnerability in the NFS server (nfsd) subsystem. The flaw involves an initialization ordering bug in the Duplicate Reply Cache (DRC) hash table. The shrinker_register() call executes before the INIT_LIST_HEAD loop and the drc_hashsize store, creating a race window where the shrinker can observe partial state.
On weakly-ordered architectures such as arm64 and ppc, a shrinker scan can read drc_hashsize before the bucket list heads are initialized. This triggers a NULL pointer dereference in the DRC shrinker callback, leading to a kernel crash.
Critical Impact
A memory shrinker invocation during NFS server initialization can cause a kernel NULL pointer dereference on weakly-ordered architectures, resulting in a system crash and denial of service.
Affected Products
- Linux kernel versions containing the pre-patch nfsd DRC initialization sequence
- Systems running on weakly-ordered architectures (arm64, ppc)
- NFS server deployments using the kernel-mode nfsd module
Discovery Timeline
- 2026-09-24 - CVE CVE-2026-93211 published to NVD
- 2026-09-24 - Last updated in NVD database
Technical Details for CVE-2026-93211
Vulnerability Analysis
The vulnerability resides in the NFS server Duplicate Reply Cache (DRC) initialization path. The DRC caches responses to non-idempotent NFS requests to protect against duplicate client retransmissions. A memory shrinker is registered to reclaim DRC entries under memory pressure.
The initialization function invokes shrinker_register() before populating the DRC hash table bucket list heads and before storing the final drc_hashsize value. A concurrent shrinker scan triggered by the memory management subsystem can observe drc_hashsize as non-zero while the bucket list heads remain uninitialized. The shrinker callback then dereferences uninitialized bucket pointers, causing a NULL pointer dereference [CWE-476].
This is a classic publication ordering bug. The visibility of drc_hashsize to other CPUs must be deferred until all dependent state is fully initialized.
Root Cause
The root cause is incorrect ordering of initialization operations relative to the shrinker registration. On strongly-ordered architectures like x86_64, the stores may happen to complete in program order and mask the race. On weakly-ordered architectures such as arm64 and ppc, the CPU and compiler are free to reorder stores, allowing the shrinker to see partial state.
Attack Vector
The condition requires memory pressure during the narrow initialization window of the nfsd subsystem. An attacker with local access on an affected architecture could potentially induce memory pressure during NFS server startup to trigger the NULL dereference. The practical attack surface is limited to systems where nfsd is being initialized under adversarial conditions.
The vulnerability manifests as a kernel oops in the shrinker callback path. See the Linux Kernel Commit b57bd8c for the specific code changes.
Detection Methods for CVE-2026-93211
Indicators of Compromise
- Kernel oops or panic messages referencing nfsd and shrinker callback functions in dmesg or /var/log/kern.log
- NULL pointer dereference backtraces in DRC-related functions during NFS server startup
- Unexpected nfsd service failures on arm64 or ppc systems coinciding with memory pressure events
Detection Strategies
- Audit running kernel versions against the fixed commits (431c70c, b0c5893, b57bd8c, f060f43) to identify unpatched systems
- Monitor kernel ring buffer output for NULL pointer dereference signatures involving DRC shrinker paths
- Correlate NFS server restarts with concurrent memory reclamation events in system telemetry
Monitoring Recommendations
- Collect kernel crash dumps using kdump to capture backtraces when the issue manifests
- Track nfsd service availability and restart counts across the Linux fleet
- Centralize dmesg output into a security data lake to pattern-match kernel oops signatures across hosts
How to Mitigate CVE-2026-93211
Immediate Actions Required
- Identify Linux hosts running nfsd on weakly-ordered architectures (arm64, ppc) and prioritize them for patching
- Apply the stable kernel update containing the fix commits as soon as vendor packages become available
- Where patching is delayed, consider disabling the nfsd kernel module on affected architectures until a fixed kernel is deployed
Patch Information
The fix moves the bucket list head initialization and the drc_hashsize store before the shrinker_register() call. This ensures the hash table is fully initialized before becoming visible to the shrinker subsystem. The upstream commits are available at Linux Kernel Commit 431c70c, Linux Kernel Commit b0c5893, Linux Kernel Commit b57bd8c, and Linux Kernel Commit f060f43.
Workarounds
- Avoid loading the nfsd module during periods of high memory pressure on arm64 and ppc hosts
- Pre-load nfsd early in the boot process before user workloads create memory contention
- Migrate NFS server workloads to x86_64 hosts temporarily if feasible, as strongly-ordered memory models reduce exposure to the race
Disclaimer: This content was generated using AI. While we strive for accuracy, please verify critical information with official sources.
Sources: National Vulnerability Database, CISA KEV, Exploit-DB

