On Tue, Dec 06, 2022 at 12:43:51PM +0100, Sebastian Andrzej Siewior wrote: > > > Before that, it did cmpxchg() which should be fine. > > > > > > Regarding mark_rt_mutex_waiters(). Isn't acquire semantic required in > > > order for the lock-owner not perform the fastpath but go to the slowpath > > > instead? > > > > > > > Good spot, it does. While the most straight-forward solution is to use > > cmpxchg_acquire, I think it is overkill because it could incur back-to-back > > ACQUIRE operations in the event of contention. There could be a smp_wmb > > after the cmpxchg_relaxed but that impacts all arches and a non-paired > > smp_wmb is generally frowned upon. > > but in general, it should succeed on the first iteration. It can only > fail (and retry) if the owner was able to unlock it first. A second > locker will spin on the wait_lock so. > Sure, generally it would be fine but it also costs us nothing to avoid additional overhead in the contended case. The pattern of atomic_relaxed+smp_mb__after_atomic is unusual but I think the comment is sufficient to explain why it's structured like that. -- Mel Gorman SUSE Labs
