Re: [PATCH v5 09/14] memcg: kmem accounting lifecycle management
From: Michal Hocko <hidden>
Date: 2012-10-18 06:14:58
Also in:
linux-mm, lkml
From: Michal Hocko <hidden>
Date: 2012-10-18 06:14:58
Also in:
linux-mm, lkml
On Wed 17-10-12 16:28:38, David Rientjes wrote:
On Tue, 16 Oct 2012, Glauber Costa wrote:
[...]
quoted
+ +static void memcg_kmem_mark_dead(struct mem_cgroup *memcg) +{ + if (test_bit(KMEM_ACCOUNTED_ACTIVE, &memcg->kmem_accounted)) + set_bit(KMEM_ACCOUNTED_DEAD, &memcg->kmem_accounted); +}The set_bit() doesn't happen atomically with the test_bit(), what synchronization is required for this?
The group has to be active in order to become dead so the ordering is natural and you do not need to test&set atomicaly. Also once a group becomes active it is always marked that way until it goes away. -- Michal Hocko SUSE Labs