jump_label: Fix the fix, brown paper bags galore

Per the example of:

  !atomic_cmpxchg(&key->enabled, 0, 1)

the inverse was written as:

  atomic_cmpxchg(&key->enabled, 1, 0)

except of course, that while !old is only true for old == 0, old is
true for everything except old == 0.

Fix it to read:

  atomic_cmpxchg(&key->enabled, 1, 0) == 1

such that only the 1->0 transition returns true and goes on to disable
the keys.

Fixes: 83ab38ef0a ("jump_label: Fix concurrency issues in static_key_slow_dec()")
Reported-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Darrick J. Wong <djwong@kernel.org>
Link: https://lkml.kernel.org/r/20240731105557.GY33588@noisy.programming.kicks-ass.net
This commit is contained in:
Peter Zijlstra 2024-07-31 12:43:21 +02:00
parent 6623b0217d
commit 224fa35520

View File

@ -236,7 +236,7 @@ void static_key_disable_cpuslocked(struct static_key *key)
} }
jump_label_lock(); jump_label_lock();
if (atomic_cmpxchg(&key->enabled, 1, 0)) if (atomic_cmpxchg(&key->enabled, 1, 0) == 1)
jump_label_update(key); jump_label_update(key);
jump_label_unlock(); jump_label_unlock();
} }
@ -289,7 +289,7 @@ static void __static_key_slow_dec_cpuslocked(struct static_key *key)
return; return;
guard(mutex)(&jump_label_mutex); guard(mutex)(&jump_label_mutex);
if (atomic_cmpxchg(&key->enabled, 1, 0)) if (atomic_cmpxchg(&key->enabled, 1, 0) == 1)
jump_label_update(key); jump_label_update(key);
else else
WARN_ON_ONCE(!static_key_slow_try_dec(key)); WARN_ON_ONCE(!static_key_slow_try_dec(key));