Re: [patch 1/5] mm: memcontrol: convert reclaim iterator to simple css refcounting

[Johannes Weiner]

On Wed, Oct 15, 2014 at 05:02:56PM +0200, Michal Hocko wrote:
> On Tue 14-10-14 12:20:33, Johannes Weiner wrote:
> > The memcg reclaim iterators use a complicated weak reference scheme to
> > prevent pinning cgroups indefinitely in the absence of memory pressure.
> > 
> > However, during the ongoing cgroup core rework, css lifetime has been
> > decoupled such that a pinned css no longer interferes with removal of
> > the user-visible cgroup, and all this complexity is now unnecessary.
> >
> > Signed-off-by: Johannes Weiner <hannes@xxxxxxxxxxx>
> > ---
> >  mm/memcontrol.c | 250 +++++++++++++++++---------------------------------------
> >  1 file changed, 76 insertions(+), 174 deletions(-)
> > 
> > diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> > index b62972c80055..67dabe8b0aa6 100644
> > --- a/mm/memcontrol.c
> > +++ b/mm/memcontrol.c
> [...]
> > +		do {
> > +			pos = ACCESS_ONCE(iter->position);
> > +			/*
> > +			 * A racing update may change the position and
> > +			 * put the last reference, hence css_tryget(),
> > +			 * or retry to see the updated position.
> > +			 */
> > +		} while (pos && !css_tryget(&pos->css));
> > +	}
> [...]
> > +	if (reclaim) {
> > +		if (cmpxchg(&iter->position, pos, memcg) == pos && memcg)
> > +			css_get(&memcg->css);
> > +
> > +		if (pos)
> > +			css_put(&pos->css);
> 
> This looks like a reference leak. css_put pairs with the above
> css_tryget but no css_put pairs with css_get for the cached one. We
> need:
> ---
> From 2810937ec6c16afc0bf924e761ff8305bd478a42 Mon Sep 17 00:00:00 2001
> From: Michal Hocko <mhocko@xxxxxxx>
> Date: Wed, 15 Oct 2014 16:26:22 +0200
> Subject: [PATCH] 
>  mm-memcontrol-convert-reclaim-iterator-to-simple-css-refcounting-fix
> 
> Make sure that the cached reference is always released.
> 
> Signed-off-by: Michal Hocko <mhocko@xxxxxxx>

You're right, thanks for catching that.

Acked-by: Johannes Weiner <hannes@xxxxxxxxxxx>
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/