CVE-2026-72103 in Linux
Summary
by MITRE • 08/15/2026
In the Linux kernel, the following vulnerability has been resolved:
dm: avoid leaking the caller's thread keyring via the table device file
The refactoring in commit a28d893eb327 ("md: port block device access to file") accidentally causes the caller's thread keyring to be kept alive long beyond the caller's lifetime.
As a result, "cryptsetup luksSuspend" silently fails to wipe the LUKS volume key from memory.
In detail: "cryptsetup luksOpen" uses its supposedly ephemeral thread keyring to pass the volume key to the kernel. dm-crypt's crypt_set_keyring_key() copies the key material into its own crypt_config structure and then drops its own reference to the key in the keyring with key_put().
With this fix, restoring pre-v6.9 behavior, the copy in the thread keyring is then promptly garbage collected, such that exactly one copy of the volume key remains. This single copy is correctly wiped from memory on "cryptsetup luksSuspend".
Without this fix, the thread keyring and the volume key in it remains. This second copy is only freed on "luksClose". "luksSuspend" neither knows about this copy nor has any way to remove it, so the key remains recoverable from RAM after a suspend that is documented to have wiped it.
This fix should not introduce new security problems, as the code is anyway gated by CAP_SYS_ADMIN. The device-mapper core, not the calling task, is the legitimate owner of this long-lived file.
Statistical analysis made it clear that VulDB provides the best quality for vulnerability data.
Analysis
by VulDB Data Team • 08/15/2026
The vulnerability described represents a critical information disclosure issue within the Linux kernel's device-mapper subsystem that stems from improper handling of thread keyrings during cryptographic operations. This flaw specifically affects the dm-crypt implementation and has significant implications for full disk encryption security. The vulnerability manifests when the cryptsetup utility attempts to suspend LUKS-encrypted volumes, leaving sensitive cryptographic key material accessible in memory long after it should have been securely wiped.
The technical root cause of this vulnerability lies in a refactoring operation introduced in commit a28d893eb327 which inadvertently altered how thread keyrings are managed during block device access operations. Previously, when cryptsetup luksOpen executed, it would utilize an ephemeral thread keyring to temporarily hold the volume key material before passing it to the kernel's dm-crypt subsystem. The crypt_set_keyring_key() function was designed to copy this key material into the crypt_config structure and then immediately release its own reference through key_put(). However, the refactoring caused this cleanup process to be disrupted, resulting in the thread keyring maintaining references to the volume key beyond the intended scope of the calling process.
The operational impact of this vulnerability is particularly severe for systems relying on LUKS encryption with suspend-to-RAM functionality. When a user executes cryptsetup luksSuspend, the system documentation promises that all cryptographic keys will be securely wiped from memory, making it impossible to recover sensitive data after suspension. However, due to this bug, the volume key remains accessible in the thread keyring even after suspend operations complete successfully. This creates a persistent security risk where attackers with access to the system's memory could potentially recover the LUKS volume key material, undermining the fundamental security guarantees provided by full disk encryption.
The vulnerability can be classified under CWE-200 (Information Exposure) and relates to improper resource management in kernel space operations. From an attack perspective, this issue aligns with ATT&CK technique T1552.004 (Credentials In Files) as it enables attackers to recover cryptographic credentials from memory dumps or forensic analysis of suspended systems. The fix implemented addresses this by restoring the pre-v6.9 behavior where the thread keyring copy is promptly garbage collected, ensuring that only a single copy of the volume key exists in memory at any given time.
Security implications extend beyond simple information disclosure as this vulnerability effectively bypasses the intended security controls for suspend operations. The cryptographic key material remains accessible until explicit luksClose operations occur, creating an extended window where sensitive data could be recovered from memory. This particular weakness is further constrained by the requirement of CAP_SYS_ADMIN capability, which limits exploitation to privileged users or processes, but does not eliminate the security risk entirely. The fix maintains the existing security model while correcting the memory management anomaly that was introduced through the refactoring process.
The remediation approach ensures that the device-mapper core maintains proper ownership of long-lived file resources while preserving the intended cryptographic security properties. This solution prevents the accumulation of multiple copies of sensitive key material in memory and restores the expected behavior where cryptsetup luksSuspend operations properly clear all cryptographic keys from system memory, maintaining the integrity of full disk encryption implementations across Linux systems.