On 4/15/21 10:25 AM, Ali Saidi wrote:
While this code is executed with the wait_lock held, a reader can acquire the lock without holding wait_lock. The writer side loops checking the value with the atomic_cond_read_acquire(), but only truly acquires the lock when the compare-and-exchange is completed successfully which isn’t ordered. The other atomic operations from this point are release-ordered and thus reads after the lock acquisition can be completed before the lock is truly acquired which violates the guarantees the lock should be making.Fixes: b519b56e378ee ("locking/qrwlock: Use atomic_cond_read_acquire() when spinning in qrwloc") Signed-off-by: Ali Saidi <alisa...@amazon.com> Cc: sta...@vger.kernel.org --- kernel/locking/qrwlock.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/kernel/locking/qrwlock.c b/kernel/locking/qrwlock.c index 4786dd271b45..10770f6ac4d9 100644 --- a/kernel/locking/qrwlock.c +++ b/kernel/locking/qrwlock.c @@ -73,8 +73,8 @@ void queued_write_lock_slowpath(struct qrwlock *lock)/* When no more readers or writers, set the locked flag */do { - atomic_cond_read_acquire(&lock->cnts, VAL == _QW_WAITING); - } while (atomic_cmpxchg_relaxed(&lock->cnts, _QW_WAITING, + atomic_cond_read_relaxed(&lock->cnts, VAL == _QW_WAITING); + } while (atomic_cmpxchg_acquire(&lock->cnts, _QW_WAITING, _QW_LOCKED) != _QW_WAITING); unlock: arch_spin_unlock(&lock->wait_lock);
I think the original code isn't wrong. The read_acquire provides the acquire barrier for cmpxchg. Because of conditional dependency, the wait_lock unlock won't happen until the cmpxchg succeeds. Without doing a read_acquire, there may be a higher likelihood that the cmpxchg may fail.
Anyway, I will let Will or Peter chime in on this as I am not as proficient as them on this topic.
Cheers, Longman
