Executive brief
The Linux kernel's AFS filesystem implementation has a flaw in how it initializes and manages work queue structures for file locking operations. This can cause the kernel to detect corrupted state and trigger warnings when file locks are used, potentially affecting system stability and causing intermittent failures in applications that rely on file locking.
Technical details
The vulnerability is a work_struct initialization flaw in the AFS filesystem module. The lock_work delayed work structure was initialized only once in the slab allocator's init function, but the kernel's DEBUG_OBJECTS checking requires reinitializaton for reused objects. When an inode is allocated and reused, the lock_work structure enters an invalid state, causing debug_object_activate() to fail when the work is queued. The fix reinitializes lock_work during inode allocation in afs_alloc_inode() and flushes pending work during inode eviction in afs_evict_inode(). This affects all Linux kernels with the AFS filesystem module enabled.
Affected products
- Linux Linux kernel all versions with AFS filesystem support
Timeline
- 2026-08-15: disclosed
- 2026-07-01: patched: Patch available in stable kernel versions
- 2026-06-22: other: Fix authored by David Howells