[PATCH] mm: memcontrol: flush percpu vmevents before releasing memcg

[PATCH] mm: memcontrol: flush percpu vmevents before releasing memcg

[Roman Gushchin]

Similar to vmstats, percpu caching of local vmevents leads to an
accumulation of errors on non-leaf levels. This happens because
some leftovers may remain in percpu caches, so that they are
never propagated up by the cgroup tree and just disappear into
nonexistence with on releasing of the memory cgroup.

To fix this issue let's accumulate and propagate percpu vmevents
values before releasing the memory cgroup similar to what we're
doing with vmstats.

Since on cpu hotplug we do flush percpu vmstats anyway, we can
iterate only over online cpus.

Fixes: 42a300353577 ("mm: memcontrol: fix recursive statistics correctness & scalabilty")
Signed-off-by: Roman Gushchin <guro@fb.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
---
 mm/memcontrol.c | 22 +++++++++++++++++++++-
 1 file changed, 21 insertions(+), 1 deletion(-)

diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 6d2427abcc0c..249187907339 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -3459,6 +3459,25 @@ static void memcg_flush_percpu_vmstats(struct mem_cgroup *memcg, bool slab_only)
 	}
 }
 
+static void memcg_flush_percpu_vmevents(struct mem_cgroup *memcg)
+{
+	unsigned long events[NR_VM_EVENT_ITEMS];
+	struct mem_cgroup *mi;
+	int cpu, i;
+
+	for (i = 0; i < NR_VM_EVENT_ITEMS; i++)
+		events[i] = 0;
+
+	for_each_online_cpu(cpu)
+		for (i = 0; i < NR_VM_EVENT_ITEMS; i++)
+			events[i] += raw_cpu_read(
+				memcg->vmstats_percpu->events[i]);
+
+	for (mi = memcg; mi; mi = parent_mem_cgroup(mi))
+		for (i = 0; i < NR_VM_EVENT_ITEMS; i++)
+			atomic_long_add(events[i], &mi->vmevents[i]);
+}
+
 static void memcg_offline_kmem(struct mem_cgroup *memcg)
 {
 	struct cgroup_subsys_state *css;
@@ -4860,10 +4879,11 @@ static void __mem_cgroup_free(struct mem_cgroup *memcg)
 	int node;
 
 	/*
-	 * Flush percpu vmstats to guarantee the value correctness
+	 * Flush percpu vmstats and vmevents to guarantee the value correctness
 	 * on parent's and all ancestor levels.
 	 */
 	memcg_flush_percpu_vmstats(memcg, false);
+	memcg_flush_percpu_vmevents(memcg);
 	for_each_node(node)
 		free_mem_cgroup_per_node_info(memcg, node);
 	free_percpu(memcg->vmstats_percpu);
-- 
2.21.0

Re: [PATCH] mm: memcontrol: flush percpu vmevents before releasing memcg

[Roman Gushchin]

On Mon, Aug 12, 2019 at 04:37:54PM -0700, Roman Gushchin wrote:
> Similar to vmstats, percpu caching of local vmevents leads to an
> accumulation of errors on non-leaf levels. This happens because
> some leftovers may remain in percpu caches, so that they are
> never propagated up by the cgroup tree and just disappear into
> nonexistence with on releasing of the memory cgroup.
> 
> To fix this issue let's accumulate and propagate percpu vmevents
> values before releasing the memory cgroup similar to what we're
> doing with vmstats.
> 
> Since on cpu hotplug we do flush percpu vmstats anyway, we can
> iterate only over online cpus.

Just to clarify: this patch should be placed on top of two other
patches, which I sent a bit earlier today:

1) mm: memcontrol: flush percpu slab vmstats on kmem offlining
2) mm: memcontrol: flush percpu vmstats before releasing memcg

Sorry for the inconvenience, I forgot about vmevents during
working on the final version, and remembered too late.

Thanks!

Re: [PATCH] mm: memcontrol: flush percpu vmevents before releasing memcg

[Andrew Morton]

On Mon, 12 Aug 2019 16:37:54 -0700 Roman Gushchin <guro@fb.com> wrote:

> Similar to vmstats, percpu caching of local vmevents leads to an
> accumulation of errors on non-leaf levels. This happens because
> some leftovers may remain in percpu caches, so that they are
> never propagated up by the cgroup tree and just disappear into
> nonexistence with on releasing of the memory cgroup.
> 
> To fix this issue let's accumulate and propagate percpu vmevents
> values before releasing the memory cgroup similar to what we're
> doing with vmstats.
> 
> Since on cpu hotplug we do flush percpu vmstats anyway, we can
> iterate only over online cpus.
> 
> Fixes: 42a300353577 ("mm: memcontrol: fix recursive statistics correctness & scalabilty")

No cc:stable?

Re: [PATCH] mm: memcontrol: flush percpu vmevents before releasing memcg

[Roman Gushchin]

On Tue, Aug 13, 2019 at 02:31:17PM -0700, Andrew Morton wrote:
> On Mon, 12 Aug 2019 16:37:54 -0700 Roman Gushchin <guro@fb.com> wrote:
> 
> > Similar to vmstats, percpu caching of local vmevents leads to an
> > accumulation of errors on non-leaf levels. This happens because
> > some leftovers may remain in percpu caches, so that they are
> > never propagated up by the cgroup tree and just disappear into
> > nonexistence with on releasing of the memory cgroup.
> > 
> > To fix this issue let's accumulate and propagate percpu vmevents
> > values before releasing the memory cgroup similar to what we're
> > doing with vmstats.
> > 
> > Since on cpu hotplug we do flush percpu vmstats anyway, we can
> > iterate only over online cpus.
> > 
> > Fixes: 42a300353577 ("mm: memcontrol: fix recursive statistics correctness & scalabilty")
> 
> No cc:stable?
> 

Here too, cc:stable is definitely missing. Adding now. Thanks!

Re: [PATCH] mm: memcontrol: flush percpu vmevents before releasing memcg

[Michal Hocko]

On Mon 12-08-19 16:37:54, Roman Gushchin wrote:
> Similar to vmstats, percpu caching of local vmevents leads to an
> accumulation of errors on non-leaf levels. This happens because
> some leftovers may remain in percpu caches, so that they are
> never propagated up by the cgroup tree and just disappear into
> nonexistence with on releasing of the memory cgroup.
> 
> To fix this issue let's accumulate and propagate percpu vmevents
> values before releasing the memory cgroup similar to what we're
> doing with vmstats.
> 
> Since on cpu hotplug we do flush percpu vmstats anyway, we can
> iterate only over online cpus.
> 
> Fixes: 42a300353577 ("mm: memcontrol: fix recursive statistics correctness & scalabilty")
> Signed-off-by: Roman Gushchin <guro@fb.com>
> Cc: Johannes Weiner <hannes@cmpxchg.org>

Acked-by: Michal Hocko <mhocko@suse.com>

> ---
>  mm/memcontrol.c | 22 +++++++++++++++++++++-
>  1 file changed, 21 insertions(+), 1 deletion(-)
> 
> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> index 6d2427abcc0c..249187907339 100644
> --- a/mm/memcontrol.c
> +++ b/mm/memcontrol.c
> @@ -3459,6 +3459,25 @@ static void memcg_flush_percpu_vmstats(struct mem_cgroup *memcg, bool slab_only)
>  	}
>  }
>  
> +static void memcg_flush_percpu_vmevents(struct mem_cgroup *memcg)
> +{
> +	unsigned long events[NR_VM_EVENT_ITEMS];
> +	struct mem_cgroup *mi;
> +	int cpu, i;
> +
> +	for (i = 0; i < NR_VM_EVENT_ITEMS; i++)
> +		events[i] = 0;
> +
> +	for_each_online_cpu(cpu)
> +		for (i = 0; i < NR_VM_EVENT_ITEMS; i++)
> +			events[i] += raw_cpu_read(
> +				memcg->vmstats_percpu->events[i]);
> +
> +	for (mi = memcg; mi; mi = parent_mem_cgroup(mi))
> +		for (i = 0; i < NR_VM_EVENT_ITEMS; i++)
> +			atomic_long_add(events[i], &mi->vmevents[i]);
> +}
> +
>  static void memcg_offline_kmem(struct mem_cgroup *memcg)
>  {
>  	struct cgroup_subsys_state *css;
> @@ -4860,10 +4879,11 @@ static void __mem_cgroup_free(struct mem_cgroup *memcg)
>  	int node;
>  
>  	/*
> -	 * Flush percpu vmstats to guarantee the value correctness
> +	 * Flush percpu vmstats and vmevents to guarantee the value correctness
>  	 * on parent's and all ancestor levels.
>  	 */
>  	memcg_flush_percpu_vmstats(memcg, false);
> +	memcg_flush_percpu_vmevents(memcg);
>  	for_each_node(node)
>  		free_mem_cgroup_per_node_info(memcg, node);
>  	free_percpu(memcg->vmstats_percpu);
> -- 
> 2.21.0

-- 
Michal Hocko
SUSE Labs