From mboxrd@z Thu Jan 1 00:00:00 1970 From: Glauber Costa Subject: Re: [PATCH v4 09/14] memcg: kmem accounting lifecycle management Date: Fri, 12 Oct 2012 11:47:17 +0400 Message-ID: <5077CB05.907@parallels.com> References: <1349690780-15988-1-git-send-email-glommer@parallels.com> <1349690780-15988-10-git-send-email-glommer@parallels.com> <20121011131143.GF29295@dhcp22.suse.cz> Mime-Version: 1.0 Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20121011131143.GF29295-2MMpYkNvuYDjFM9bn6wA6Q@public.gmane.org> Sender: cgroups-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org List-ID: Content-Type: text/plain; charset="us-ascii" To: Michal Hocko Cc: linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, linux-mm-Bw31MaZKKs3YtjvyW6yDsg@public.gmane.org, Andrew Morton , Mel Gorman , Suleiman Souhlal , Tejun Heo , cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, kamezawa.hiroyu-+CUm20s59erQFUHtdCDX3A@public.gmane.org, Johannes Weiner , Greg Thelen , devel-GEFAQzZX7r8dnm+yROfE0A@public.gmane.org, Frederic Weisbecker , Christoph Lameter , Pekka Enberg On 10/11/2012 05:11 PM, Michal Hocko wrote: > On Mon 08-10-12 14:06:15, Glauber Costa wrote: >> Because kmem charges can outlive the cgroup, we need to make sure that >> we won't free the memcg structure while charges are still in flight. >> For reviewing simplicity, the charge functions will issue >> mem_cgroup_get() at every charge, and mem_cgroup_put() at every >> uncharge. >> >> This can get expensive, however, and we can do better. mem_cgroup_get() >> only really needs to be issued once: when the first limit is set. In the >> same spirit, we only need to issue mem_cgroup_put() when the last charge >> is gone. >> >> We'll need an extra bit in kmem_accounted for that: KMEM_ACCOUNTED_DEAD. >> it will be set when the cgroup dies, if there are charges in the group. >> If there aren't, we can proceed right away. >> >> Our uncharge function will have to test that bit every time the charges >> drop to 0. Because that is not the likely output of >> res_counter_uncharge, this should not impose a big hit on us: it is >> certainly much better than a reference count decrease at every >> operation. >> >> [ v3: merged all lifecycle related patches in one ] >> >> Signed-off-by: Glauber Costa >> CC: Kamezawa Hiroyuki >> CC: Christoph Lameter >> CC: Pekka Enberg >> CC: Michal Hocko >> CC: Johannes Weiner >> CC: Suleiman Souhlal > > OK, I like the optimization. I have just one comment to the > memcg_kmem_dead naming but other than that > > Acked-by: Michal Hocko > > [...] >> +static bool memcg_kmem_dead(struct mem_cgroup *memcg) > > The name is tricky because it doesn't tell you that it clears the flag > which made me scratch my head when reading comment in kmem_cgroup_destroy > memcg_kmem_finally_kill_that_bastard() ?