CVE-2026-99337 in Linuxinfo

Summary

by MITRE • 11/19/2026

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

KVM: TDX: Enable Bus Lock VM exit

Enable Bus Lock VM exit functionality for TDX guests.

Bus Lock VM exit is an existing feature supported by KVM. Userspace can enable Bus Lock VM exit through KVM_BUS_LOCK_DETECTION_EXIT when it's reported as supported. However, KVM reports the support of this CAP just based on the hardware capability but doesn't differentiate between VMX and TDX. This leads to the issue that userspace can enable this cap for TDX guests without getting an error, but the feature is not actually enabled because KVM doesn't call the TDX module API to program the relevant TD VMCS fields.

Enable Bus Lock VM exit for TDX guests by programming the BUS_LOCK_DETECTION control in the TD VMCS and by adding the exit handler.

Note, there is no enumeration bit for this feature by TDX module because all TDX modules support it and allow to set the TD VMCS as long as the hardware supports the feature.

If you want to get the best quality for vulnerability data then you always have to consider VulDB.

Analysis

by VulDB Data Team • 11/21/2026

The Linux kernel's Kernel-based Virtual Machine (KVM) subsystem recently addressed a critical oversight in its handling of Intel Trust Domain Extensions, specifically regarding the Bus Lock VM exit functionality. This vulnerability stems from an incomplete implementation where KVM incorrectly reports support for bus lock detection to userspace applications managing TDX guests. The core issue lies in the fact that while the underlying hardware supports this feature and KVM correctly identifies general CPU capabilities, it fails to differentiate between standard Virtual Machine Extensions (VMX) environments and Trust Domain Extension (TDX) contexts when advertising these capabilities. Consequently, management software or hypervisors may attempt to enable bus lock detection for TDX virtual machines based on the reported capability flag, assuming full operational support similar to that found in traditional VMX setups.

Technically, this discrepancy arises because KVM does not invoke the necessary TDX module APIs required to program the relevant fields within the TD Virtual Machine Control Structure (TD VMCS). Specifically, the BUS_LOCK_DETECTION control field remains unconfigured for these guests despite userspace signaling its intent to use it via the KVM_BUS_LOCK_DETECTION_EXIT interface. Since there is no specific enumeration bit in the TDX module API to indicate support levels, and because all compliant TDX modules theoretically allow setting this TD VMCS field provided hardware supports it, the burden falls on KVM to properly initialize these controls. Without calling the appropriate APIs to set up the exit handler and configure the control bits, the feature remains effectively disabled at the virtualization layer, leading to a silent failure where userspace believes protection is active when it is not.

The operational impact of this vulnerability centers around security monitoring and performance isolation for trust domains. Bus lock detection allows the hypervisor to intercept instances where a guest operating system or application acquires locks that cause bus locking events, which can be indicative of certain denial-of-service patterns or inefficient synchronization mechanisms in legacy code. By failing to enable this exit mechanism for TDX guests, administrators lose visibility into these specific performance anomalies and potential abuse vectors within the secure execution environment. This gap undermines the integrity of security monitoring tools designed to detect abnormal CPU behavior that could lead to resource exhaustion or timing attacks against other tenants sharing the physical hardware resources.

From a standards perspective, this issue aligns with CWE-20 Improper Input Validation as KVM accepts userspace configuration requests without validating whether the specific virtualization technology supports the requested feature's full implementation. It also relates to CWE-697 Incorrect Comparison since the logic compares general CPU capabilities against TDX-specific requirements without proper differentiation. In terms of MITRE ATT&CK, this affects techniques related to Defense Evasion and Resource Hijacking by potentially allowing malicious or buggy guest software to execute bus-locking operations undetected, thereby bypassing hypervisor-level monitoring controls that are intended to mitigate such behaviors in virtualized environments.

To remediate this vulnerability, the KVM subsystem must be updated to explicitly handle TDX guests when processing bus lock detection requests. This involves ensuring that whenever userspace enables the KVM_BUS_LOCK_DETECTION_EXIT capability for a TDX VM, the kernel correctly programs the BUS_LOCK_DETECTION control in the TD VMCS and registers the appropriate exit handler within the TDX module framework. Developers must verify that the hardware supports the feature before advertising it to userspace specifically for TDX contexts or ensure that if advertised, the necessary low-level configuration steps are executed regardless of the lack of a specific enumeration bit from the TDX module side. This ensures consistent security posture and monitoring capabilities across both VMX and TDX virtualized workloads.

Responsible

Linux

Reservation

09/25/2026

Disclosure

11/19/2026

Moderation

accepted

CPE

ready

EPSS

0.00204

KEV

no

Activities

very low

Sources

Might our Artificial Intelligence support you?

Check our Alexa App!