Messages in this thread | | | Date | Thu, 30 Jan 2014 12:29:06 -0500 | From | Johannes Weiner <> | Subject | Re: [RFC 4/5] memcg: make sure that memcg is not offline when charging |
| |
On Tue, Dec 17, 2013 at 04:45:29PM +0100, Michal Hocko wrote: > The current charge path might race with memcg offlining because holding > css reference doesn't stop css offline. As a result res counter might be > charged after mem_cgroup_reparent_charges (called from memcg css_offline > callback) and so the charge would never be freed. This has been worked > around by 96f1c58d8534 (mm: memcg: fix race condition between memcg > teardown and swapin) which tries to catch such a leaked charges later > during css_free. It is more optimal to heal this race in the long term > though.
We already deal with the race, so IMO the only outstanding improvement is to take advantage of the teardown synchronization provided by the cgroup core and get rid of our one-liner workaround in .css_free.
> In order to make this raceless we would need to hold rcu_read_lock since > css_tryget until res_counter_charge. This is not so easy unfortunately > because mem_cgroup_do_charge might sleep so we would need to do drop rcu > lock and do css_tryget tricks after each reclaim.
Yes, why not?
> This patch addresses the issue by introducing memcg->offline flag > which is set from mem_cgroup_css_offline callback before the pages are > reparented. mem_cgroup_do_charge checks the flag before res_counter > is charged inside rcu read section. mem_cgroup_css_offline uses > synchronize_rcu to let all preceding chargers finish while all the new > ones will see the group offline already and back out. > > Callers are then updated to retry with a new memcg which is fallback to > mem_cgroup_from_task(current). > > The only exception is mem_cgroup_do_precharge which should never see > this race because it is called from cgroup {can_}attach callbacks and so > the whole cgroup cannot go away. > > Signed-off-by: Michal Hocko <mhocko@suse.cz> > --- > mm/memcontrol.c | 58 ++++++++++++++++++++++++++++++++++++++++++++++++++++++--- > 1 file changed, 55 insertions(+), 3 deletions(-)
That makes no sense to me. It's a lateral move in functionality and cgroup integration, but more complicated.
| |