Re: [PATCH v3 7/7] locking/rtmutex: Acquire the hb lock via trylock after wait-proxylock.

[Peter Zijlstra]

On Tue, Sep 12, 2023 at 01:17:11PM +0200, Sebastian Andrzej Siewior wrote:
> On 2023-09-11 16:11:35 [+0200], Peter Zijlstra wrote:
> > Aside from this being just plain gross, this also destroys determinism
> > of futex_pi, which completely defeats the purpose of the whole thing.
> 
> No objections here.
> 
> > Now.. the reason we need hb->lock at this point is to avoid the
> > wake_futex_pi() -EAGAIN case.
> > 
> > This happens when futex_top_waiter() and rt_mutex_top_waiter() state
> > becomes inconsistent. The current rules are such that this inconsistency
> > will not be observed.
> > 
> > Notably the case that needs to be avoided is where futex_lock_pi() and
> > futex_unlock_pi() interleave such that unlock will fail to observe a new
> > waiter.
> > 
> > *However* the case at hand is where a waiter is leaving, in this case
> > what happens is that we'll fail to observe a waiter that is already
> > gone, which is harmless afaict.
> 
> Sounds harmless. I sure something will be pop up in a few years and we
> will look back this ;)

Oh absolutely, I already hate this... It's dangerous as heck, because
if we ever do trip the race the other way around it will silently
misbehave.

However, compared to the other hack ... :-(