On Mon, Jan 09, 2012 at 03:04:04PM +0100, Johannes Weiner wrote: > On Fri, Jan 06, 2012 at 10:57:50PM +0200, Kirill A. Shutemov wrote: > > From: "Kirill A. Shutemov" <kirill@xxxxxxxxxxxxx> > > > > action != CPU_DEAD || action != CPU_DEAD_FROZEN is always true. > > > > Signed-off-by: Kirill A. Shutemov <kirill@xxxxxxxxxxxxx> > > Cc: <stable@xxxxxxxxxx> > > I think you don't need to actually CC stable via email. If you > include that tag, they will pick it up once the patch hits mainline. I don't think it's a problem for stable@. > > The changelog is too terse, doubly so for a patch that should go into > stable. How is the code supposed to work? What are the consequences > of the bug? Is it enough? --- >From fe1c2f2dc1abf63cce12017e2d9031cf67f0a161 Mon Sep 17 00:00:00 2001 From: "Kirill A. Shutemov" <kirill@xxxxxxxxxxxxx> Date: Sat, 24 Dec 2011 04:12:53 +0200 Subject: [PATCH 4/6] memcg: fix broken boolean expression action != CPU_DEAD || action != CPU_DEAD_FROZEN is always true. We should return at the point if CPU doesn't go away. Otherwise drain all counters and stocks from the CPU. Signed-off-by: Kirill A. Shutemov <kirill@xxxxxxxxxxxxx> Cc: <stable@xxxxxxxxxx> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@xxxxxxxxxxxxxx> Acked-by: Michal Hocko <mhocko@xxxxxxx> --- mm/memcontrol.c | 4 +++- 1 files changed, 3 insertions(+), 1 deletions(-) diff --git a/mm/memcontrol.c b/mm/memcontrol.c index 4ed6737..513ae04 100644 --- a/mm/memcontrol.c +++ b/mm/memcontrol.c @@ -2095,9 +2095,11 @@ static int __cpuinit memcg_cpu_hotplug_callback(struct notifier_block *nb, return NOTIFY_OK; } - if ((action != CPU_DEAD) || action != CPU_DEAD_FROZEN) + if (action != CPU_DEAD && action != CPU_DEAD_FROZEN) return NOTIFY_OK; + /* CPU goes away */ + for_each_mem_cgroup(iter) mem_cgroup_drain_pcp_counter(iter, cpu); -- 1.7.8.3 -- Kirill A. Shutemov _______________________________________________ Containers mailing list Containers@xxxxxxxxxxxxxxxxxxxxxxxxxx https://lists.linuxfoundation.org/mailman/listinfo/containers
