"value":"In the Linux kernel, the following vulnerability has been resolved:\n\nsmb: During unmount, ensure all cached dir instances drop their dentry\n\nThe unmount process (cifs_kill_sb() calling close_all_cached_dirs()) can\nrace with various cached directory operations, which ultimately results\nin dentries not being dropped and these kernel BUGs:\n\nBUG: Dentry ffff88814f37e358{i=1000000000080,n=/} still in use (2) [unmount of cifs cifs]\nVFS: Busy inodes after unmount of cifs (cifs)\n------------[ cut here ]------------\nkernel BUG at fs/super.c:661!\n\nThis happens when a cfid is in the process of being cleaned up when, and\nhas been removed from the cfids->entries list, including:\n\n- Receiving a lease break from the server\n- Server reconnection triggers invalidate_all_cached_dirs(), which\n removes all the cfids from the list\n- The laundromat thread decides to expire an old cfid.\n\nTo solve these problems, dropping the dentry is done in queued work done\nin a newly-added cfid_put_wq workqueue, and close_all_cached_dirs()\nflushes that workqueue after it drops all the dentries of which it's\naware. This is a global workqueue (rather than scoped to a mount), but\nthe queued work is minimal.\n\nThe final cleanup work for cleaning up a cfid is performed via work\nqueued in the serverclose_wq workqueue; this is done separate from\ndropping the dentries so that close_all_cached_dirs() doesn't block on\nany server operations.\n\nBoth of these queued works expect to invoked with a cfid reference and\na tcon reference to avoid those objects from being freed while the work\nis ongoing.\n\nWhile we're here, add proper locking to close_all_cached_dirs(), and\nlocking around the freeing of cfid->dentry."
