[PATCH] locking/lockdep: Invalidate stale class_cache entries for zapped classes
From: Eric Dumazet <edumazet@google.com>
Date: 2026-08-24 15:51:32
Also in:
lkml
Subsystem:
locking primitives, the rest · Maintainers:
Peter Zijlstra, Ingo Molnar, Will Deacon, Boqun Feng, Linus Torvalds
syzbot reported a lockdep splat hitting DEBUG_LOCKS_WARN_ON(1) in
hlock_class() due to an invalid class_idx:
WARNING: kernel/locking/lockdep.c:238 at __lock_acquire+0x382/0x2cf0 kernel/locking/lockdep.c:5203
Workqueue: wg-crypt-wg0 wg_packet_tx_worker
RIP: 0010:hlock_class kernel/locking/lockdep.c:238 [inline]
RIP: 0010:check_wait_context kernel/locking/lockdep.c:4870 [inline]
RIP: 0010:__lock_acquire+0x389/0x2cf0 kernel/locking/lockdep.c:5203
Call Trace:
<IRQ>
lock_acquire+0x106/0x350 kernel/locking/lockdep.c:5886
_raw_spin_lock+0x2e/0x40 kernel/locking/spinlock.c:173
tcp_tsq_handler+0x29/0x200 net/ipv4/tcp_output.c:1291
tcp_tsq_workfn+0x384/0x410 net/ipv4/tcp_output.c:1325
...
When a lock class is zapped (e.g. during module unload or key
unregistration), zap_class() clears the class's bit in
lock_classes_in_use and removes it from the class hash table.
However, existing lockdep_map instances embedded in data structures
may still retain a pointer to the zapped class in their class_cache[]
array.
When __lock_acquire() subsequently runs on such a lock, it finds
lock->class_cache[subclass] != NULL, skipping register_lock_class() and
assigning hlock->class_idx to the index of the zapped class. When
check_wait_context() or hlock_class() inspects the held_lock, it finds
!test_bit(class_idx, lock_classes_in_use) and warns. Furthermore, if
the zapped slot is subsequently re-allocated to an unrelated lock key,
the stale class_cache entry would erroneously match the unrelated class
(ABA issue).
Add lock_class_cache_is_valid() to validate that the cached class is
within lock_classes bounds, still allocated in lock_classes_in_use
(using uninstrumented arch_test_bit() in __always_inline context so it
is safe in noinstr contexts like match_held_lock()), and that class->key
matches the expected subkey (taking lockdep_set_subclass() overrides into
account). Also use READ_ONCE()/WRITE_ONCE() when accessing class_cache[].
If the entry is invalid or stale, fall back to register_lock_class() /
look_up_lock_class().
Fixes: a0b0fd53e1e6 ("locking/lockdep: Free lock classes that are no longer in use")
Reported-by: syzbot+2d770620059281e225a4@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/netdev/6a8c66dc.4d75e56a.c9a88.0050.GAE@google.com/T/#u (local)
Assisted-by: Gemini:gemini-3.1-pro
Signed-off-by: Eric Dumazet <edumazet@google.com>
---
kernel/locking/lockdep.c | 50 +++++++++++++++++++++++++++++++++-------
1 file changed, 42 insertions(+), 8 deletions(-)
diff --git a/kernel/locking/lockdep.c b/kernel/locking/lockdep.c
index 25d77d4a1061..763f79806bd8 100644
--- a/kernel/locking/lockdep.c
+++ b/kernel/locking/lockdep.c@@ -963,6 +963,34 @@ look_up_lock_class(const struct lockdep_map *lock, unsigned int subclass) return NULL; } +static __always_inline bool lock_class_cache_is_valid(const struct lockdep_map *lock, + const struct lock_class *class, + unsigned int subclass) +{ + unsigned int class_subclass; + + if (!class) + return false; + + if (unlikely(class < lock_classes || class >= lock_classes + MAX_LOCKDEP_KEYS)) + return false; + + if (unlikely(!arch_test_bit(class - lock_classes, lock_classes_in_use))) + return false; + + if (unlikely(!lock->key)) + return false; + + class_subclass = subclass ? subclass : class->subclass; + if (unlikely(class_subclass >= MAX_LOCKDEP_SUBCLASSES)) + return false; + + if (unlikely(READ_ONCE(class->key) != lock->key->subkeys + class_subclass)) + return false; + + return true; +} + /* * Static locks do not have their class-keys yet - for them the key is * the lock object itself. If the lock is in the per cpu area, the
@@ -1395,9 +1423,9 @@ register_lock_class(struct lockdep_map *lock, unsigned int subclass, int force) out_set_class_cache: if (!subclass || force) - lock->class_cache[0] = class; + WRITE_ONCE(lock->class_cache[0], class); else if (subclass < NR_LOCKDEP_CACHING_CLASSES) - lock->class_cache[subclass] = class; + WRITE_ONCE(lock->class_cache[subclass], class); /* * Hash collision, did we smoke some? We found a class with a matching
@@ -4957,7 +4985,7 @@ void lockdep_init_map_type(struct lockdep_map *lock, const char *name, int i; for (i = 0; i < NR_LOCKDEP_CACHING_CLASSES; i++) - lock->class_cache[i] = NULL; + WRITE_ONCE(lock->class_cache[i], NULL); #ifdef CONFIG_LOCK_STAT lock->cpu = raw_smp_processor_id();
@@ -5022,12 +5050,15 @@ EXPORT_SYMBOL_GPL(__lockdep_no_track__); void lockdep_set_lock_cmp_fn(struct lockdep_map *lock, lock_cmp_fn cmp_fn, lock_print_fn print_fn) { - struct lock_class *class = lock->class_cache[0]; + struct lock_class *class = READ_ONCE(lock->class_cache[0]); unsigned long flags; raw_local_irq_save(flags); lockdep_recursion_inc(); + if (!lock_class_cache_is_valid(lock, class, 0)) + class = NULL; + if (!class) class = register_lock_class(lock, 0, 0);
@@ -5119,8 +5150,11 @@ static int __lock_acquire(struct lockdep_map *lock, unsigned int subclass, if (DEBUG_LOCKS_WARN_ON(subclass >= MAX_LOCKDEP_SUBCLASSES)) return 0; - if (subclass < NR_LOCKDEP_CACHING_CLASSES) - class = lock->class_cache[subclass]; + if (subclass < NR_LOCKDEP_CACHING_CLASSES) { + class = READ_ONCE(lock->class_cache[subclass]); + if (!lock_class_cache_is_valid(lock, class, subclass)) + class = NULL; + } /* * Not cached? */
@@ -5323,9 +5357,9 @@ static noinstr int match_held_lock(const struct held_lock *hlock, return 1; if (hlock->references) { - const struct lock_class *class = lock->class_cache[0]; + const struct lock_class *class = READ_ONCE(lock->class_cache[0]); - if (!class) + if (!lock_class_cache_is_valid(lock, class, 0)) class = look_up_lock_class(lock, 0); /*
base-commit: 0a0d1d55dad570724bf8c7ea83409639cfb4be9b -- 2.55.0.766.g2966f0265a-goog