Re: [PATCH 3/4] Memory controller soft limit organize cgroups (v5)

[Balbir Singh <balbir@linux.vnet.ibm.com>]

* KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> [2009-03-13 14:22:28]:

> > > >  /*
> > > > + * Cgroups above their limits are maintained in a RB-Tree, independent of
> > > > + * their hierarchy representation
> > > > + */
> > > > +
> > > > +static struct rb_root mem_cgroup_soft_limit_tree;
> > > > +static DEFINE_SPINLOCK(memcg_soft_limit_tree_lock);
> > > 
> > > I have objection to this.
> > > Please don't use global spin lock.
> > 
> > We need a global data structure, per node, per zone is no good, since
> > the limits (soft limit in this case) is for the entire cgroup.
> 
> this smell the data structure is wrong.
> 
> rb-tree soring is one of efficient reclaiming technique.
> but global lock bust due to this patch's good side.
>

memory cgroup is a global data structure. Can we have the same mem
cgorup in several RB-Trees?
 
> if its updating is really rare, rcu is better?
> or couldn't you select another data structure?

RCU for RB-Trees? We'll need to RCU'fy all the links and the core
RB-Tree. RB-Trees are being used in the scheduler and for
io-scheduling, hrtimers, etc. What is your concern with the data
structure?


-- 
	Balbir

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>