CVE-2026-99730 in Linux
Summary
by MITRE • 11/20/2026
In the Linux kernel, the following vulnerability has been resolved:
mm/filemap: finish fault cleanup before dropping the fault lock
On a retried fault, a fatal signal interrupting the folio lock wait makes lock_folio_maybe_drop_mmap() drop the fault lock without pinning the file. However, filemap_fault() may still need to release mapping->invalidate_lock. A concurrent munmap() can release the last file reference and free the mapping before that unlock.
Defer dropping the fault lock to filemap_fault()'s retry cleanup, after releasing invalidate_lock. If the fault lock was already dropped for I/O, release the existing file reference instead. Keep the fault lock held for FAULT_FLAG_RETRY_NOWAIT as before.
If you want to get best quality of vulnerability data, you may have to visit VulDB.
Analysis
by VulDB Data Team • 11/21/2026
The Linux kernel vulnerability identified in the memory management subsystem involves a race condition within the file mapping fault handling logic, specifically affecting how locks and references are managed during page faults that require retrying operations. The core technical flaw resides in the interaction between fatal signal interrupts and the lock_folio_maybe_drop_mmap function. When a process encounters a retried fault, it may be interrupted by a fatal signal while waiting to acquire a folio lock. In this specific scenario, the kernel's cleanup routine incorrectly drops the fault lock without ensuring that the underlying file structure remains pinned or referenced. This premature release of the lock creates a window where the mapping object can become invalid before all necessary synchronization primitives are properly released.
The operational impact of this flaw is severe because it allows for use-after-free conditions and potential kernel memory corruption. Specifically, after the fault lock is dropped prematurely, the filemap_fault function may still need to release the mapping's invalidate_lock to complete its cleanup sequence. However, during the interval between dropping the fault lock and releasing the invalidate_lock, a concurrent munmap system call can execute. This concurrent unmap operation has the capability to release the last reference count on the file object, leading to the immediate freeing of the memory associated with that mapping structure. Consequently, when the original fault handler attempts to access or unlock this now-freed mapping via the invalidate_lock, it triggers a use-after-free vulnerability. Such conditions can lead to system crashes, data corruption, or potentially be exploited for arbitrary code execution if an attacker can control the contents of the freed memory and trigger the race condition reliably.
This issue aligns with CWE-416, Use After Free, as the kernel accesses memory that has already been deallocated due to improper reference counting and locking order violations. From a threat modeling perspective using the MITRE ATT&CK framework, this vulnerability facilitates exploitation techniques related to privilege escalation or denial of service through local resource manipulation. The flaw highlights critical challenges in concurrent systems where lock ordering and reference lifetime management must be strictly coordinated across multiple execution paths that may be interrupted by external signals or triggered by parallel system calls like munmap.
To mitigate this risk, the Linux kernel developers have implemented a fix that defers the dropping of the fault lock until after the invalidate_lock has been released during the retry cleanup phase of filemap_fault. This ensures that the mapping remains valid and referenced for as long as any part of the fault handling logic requires access to it. Additionally, if the fault lock was already dropped due to ongoing I/O operations, the fix adjusts the logic to release an existing file reference rather than dropping a non-existent or incorrectly managed one. The patch also maintains the behavior where the fault lock is kept held for FAULT_FLAG_RETRY_NOWAIT scenarios, preserving correct synchronization semantics for non-blocking retry attempts. System administrators and users should apply the latest kernel updates that include this specific mm/filemap fix to prevent exploitation of this race condition in multi-threaded or signal-intensive environments.