Re: Advice on cgroup rstat lock

From: Yosry Ahmed
Date: Wed Apr 17 2024 - 22:05:43 EST


[..]

> > > I personally don't like mem_cgroup_flush_stats_ratelimited() very
> > > much, because it is time-based (unlike memcg_vmstats_needs_flush()),
> > > and a lot of changes can happen in a very short amount of time.
> > > However, it seems like for some workloads it's a necessary evil :/
> > >
>
> Other than obj_cgroup_may_zswap(), there is no other place which really
> need very very accurate stats. IMO we should actually make ratelimited
> version the default one for all the places. Stats will always be out of
> sync for some time window even with non-ratelimited flush and I don't
> see any place where 2 second old stat would be any issue.

We disagreed about this before, and I am not trying to get you to
debate this with me again :)

I just prefer that we avoid this if possible. We have seen cases where
the 2 sec window caused issues. Not because 2 sec is a long time, but
because userspace reads the stats after an event occurs (e.g.
proactive reclaim), but gets stats from before the event.

[..]
>
> >
> >
> > With a mutex lock contention will be less obvious, as converting this to
> > a mutex avoids multiple CPUs spinning while waiting for the lock, but
> > it doesn't remove the lock contention.
> >
>
> I don't like global sleepable locks as those are source of priority
> inversion issues on highly utilized multi-tenant systems but I still
> need to see how you are handling that.

For context, this was discussed before as well in [1].

[1]https://lore.kernel.org/lkml/CALvZod441xBoXzhqLWTZ+xnqDOFkHmvrzspr9NAr+nybqXgS-A@xxxxxxxxxxxxxx/