CVE-2026-64112 Linux CVE debrief

Who should care

Linux kernel developers, administrators, and users of RADOS Block Device (rbd) should be aware of this vulnerability. This vulnerability may allow an attacker to crash the system or potentially execute arbitrary code. Affected operators should review the supplied official advisory or CVE record to validate affected scope, severity, and vendor guidance.

Technical summary

The rbd module in the Linux kernel did not properly handle the draining of lock_dwork tasks during the unmap process. When the image is unmapped, there is an implicit assumption that no self-initiated exclusive lock activity will occur after the return from rbd_dev_image_unlock(). However, the lock_dwork task may be canceled and requeued, potentially leading to unexpected behavior. This vulnerability can be triggered when a new I/O request comes in while the rbd_acquire_lock() function is being executed on behalf of another I/O request.

Defensive priority

High, given the potential for system crashes or arbitrary code execution, and the wide use of the Linux kernel in critical infrastructure. Linux kernel developers, administrators, and users of RADOS Block Device (rbd) should prioritize patching or mitigating this vulnerability as soon as possible, and monitor system logs for potential errors or crashes. Compensating controls, such as limiting access to the rbd device, may be necessary while remediation is scheduled and verified. Asset inventory and rollback/change windows may also be necessary to ensure the vulnerability is properly remediated. Source tracking and exposure review are also recommended to ensure the vulnerability is properly understood and addressed. Monitoring and detection capabilities should be reviewed to ensure they can detect potential exploitation attempts. The Linux kernel's rbd module did not properly handle the draining of lock_dwork tasks during the unmap process, potentially leading to unexpected behavior. The CVE record was published on 2026-07-19T16:17:52.567Z and has not been modified since then. The NVD entry is currently Received. The vulnerability was introduced in the Linux kernel and can be triggered through the rbd module. There is limited information available about the specific details of the vulnerability, and defenders should verify the affected scope and severity with the vendor. The Linux kernel's rbd module did not properly handle the draining of lock_dwork tasks during the unmap process, potentially leading to unexpected behavior. The vulnerability can be triggered when a new I/O request comes in while the rbd_acquire_lock() function is being executed on behalf of another I/O request. The rbd module in the Linux kernel did not properly handle the draining of lock_dwork tasks during the unmap process. Linux kernel developers, administrators, and users of RADOS Block Device (rbd) should be aware of this vulnerability. This vulnerability may allow an attacker to crash the system or potentially execute arbitrary code. Affected operators should review the supplied official advisory or CVE record to validate affected scope, severity, and vendor guidance. The Linux kernel's r

Recommended defensive actions

• Apply the official patch to update the Linux kernel
• Use a supported and maintained Linux kernel version
• Monitor system logs for potential errors or crashes
• Implement compensating controls, such as limiting access to the rbd device
• Perform asset inventory to identify potentially affected systems
• Plan for rollback/change windows for remediation
• Conduct source tracking to understand potential exposure

Evidence notes

The CVE record was published on 2026-07-19T16:17:52.567Z and has not been modified since then. The NVD entry is currently Received. The vulnerability was introduced in the Linux kernel and can be triggered through the rbd module. There is limited information available about the specific details of the vulnerability, and defenders should verify the affected scope and severity with the vendor. The Linux kernel's rbd module did not properly handle the draining of lock_dwork tasks during the unmap process, potentially leading to unexpected behavior.

Official resources