* Re: [PATCH resend] mm, memcg: fix inconsistent oom event behavior
2020-05-02 14:10 [PATCH resend] mm, memcg: fix inconsistent oom event behavior Yafang Shao
@ 2020-05-02 14:45 ` Johannes Weiner
2020-05-02 15:23 ` Chris Down
` (2 subsequent siblings)
3 siblings, 0 replies; 6+ messages in thread
From: Johannes Weiner @ 2020-05-02 14:45 UTC (permalink / raw)
To: Yafang Shao; +Cc: akpm, linux-mm, Chris Down, Michal Hocko
On Sat, May 02, 2020 at 10:10:55AM -0400, Yafang Shao wrote:
> A recent commit 9852ae3fe529 ("mm, memcg: consider subtrees in
> memory.events") changes the behavior of memcg events, which will
> consider subtrees in memory.events. But oom_kill event is a special one
> as it is used in both cgroup1 and cgroup2. In cgroup1, it is displayed
> in memory.oom_control. The file memory.oom_control is in both root memcg
> and non root memcg, that is different with memory.event as it only in
> non-root memcg. That commit is okay for cgroup2, but it is not okay for
> cgroup1 as it will cause inconsistent behavior between root memcg and
> non-root memcg.
>
> Here's an example on why this behavior is inconsistent in cgroup1.
> root memcg
> /
> memcg foo
> /
> memcg bar
>
> Suppose there's an oom_kill in memcg bar, then the oon_kill will be
>
> root memcg : memory.oom_control(oom_kill) 0
> /
> memcg foo : memory.oom_control(oom_kill) 1
> /
> memcg bar : memory.oom_control(oom_kill) 1
>
> For the non-root memcg, its memory.oom_control(oom_kill) includes its
> descendants' oom_kill, but for root memcg, it doesn't include its
> descendants' oom_kill. That means, memory.oom_control(oom_kill) has
> different meanings in different memcgs. That is inconsistent. Then the user
> has to know whether the memcg is root or not.
>
> If we can't fully support it in cgroup1, for example by adding
> memory.events.local into cgroup1 as well, then let's don't touch
> its original behavior.
>
> Fixes: 9852ae3fe529 ("mm, memcg: consider subtrees in memory.events")
> Reviewed-by: Shakeel Butt <shakeelb@google.com>
> Cc: Chris Down <chris@chrisdown.name>
> Cc: Michal Hocko <mhocko@kernel.org>
> Cc: Johannes Weiner <hannes@cmpxchg.org>
> Signed-off-by: Yafang Shao <laoar.shao@gmail.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
^ permalink raw reply [flat|nested] 6+ messages in thread* Re: [PATCH resend] mm, memcg: fix inconsistent oom event behavior
2020-05-02 14:10 [PATCH resend] mm, memcg: fix inconsistent oom event behavior Yafang Shao
2020-05-02 14:45 ` Johannes Weiner
@ 2020-05-02 15:23 ` Chris Down
2020-05-04 7:54 ` Michal Hocko
2020-05-04 23:03 ` Andrew Morton
3 siblings, 0 replies; 6+ messages in thread
From: Chris Down @ 2020-05-02 15:23 UTC (permalink / raw)
To: Yafang Shao; +Cc: akpm, linux-mm, Michal Hocko, Johannes Weiner
Yafang Shao writes:
>A recent commit 9852ae3fe529 ("mm, memcg: consider subtrees in
>memory.events") changes the behavior of memcg events, which will
>consider subtrees in memory.events. But oom_kill event is a special one
>as it is used in both cgroup1 and cgroup2. In cgroup1, it is displayed
>in memory.oom_control. The file memory.oom_control is in both root memcg
>and non root memcg, that is different with memory.event as it only in
>non-root memcg. That commit is okay for cgroup2, but it is not okay for
>cgroup1 as it will cause inconsistent behavior between root memcg and
>non-root memcg.
Thanks!
Acked-by: Chris Down <chris@chrisdown.name>
^ permalink raw reply [flat|nested] 6+ messages in thread* Re: [PATCH resend] mm, memcg: fix inconsistent oom event behavior
2020-05-02 14:10 [PATCH resend] mm, memcg: fix inconsistent oom event behavior Yafang Shao
2020-05-02 14:45 ` Johannes Weiner
2020-05-02 15:23 ` Chris Down
@ 2020-05-04 7:54 ` Michal Hocko
2020-05-04 23:03 ` Andrew Morton
3 siblings, 0 replies; 6+ messages in thread
From: Michal Hocko @ 2020-05-04 7:54 UTC (permalink / raw)
To: Yafang Shao; +Cc: akpm, linux-mm, Chris Down, Johannes Weiner
On Sat 02-05-20 10:10:55, Yafang Shao wrote:
> A recent commit 9852ae3fe529 ("mm, memcg: consider subtrees in
> memory.events") changes the behavior of memcg events, which will
> consider subtrees in memory.events. But oom_kill event is a special one
> as it is used in both cgroup1 and cgroup2. In cgroup1, it is displayed
> in memory.oom_control. The file memory.oom_control is in both root memcg
> and non root memcg, that is different with memory.event as it only in
> non-root memcg. That commit is okay for cgroup2, but it is not okay for
> cgroup1 as it will cause inconsistent behavior between root memcg and
> non-root memcg.
>
> Here's an example on why this behavior is inconsistent in cgroup1.
> root memcg
> /
> memcg foo
> /
> memcg bar
>
> Suppose there's an oom_kill in memcg bar, then the oon_kill will be
>
> root memcg : memory.oom_control(oom_kill) 0
> /
> memcg foo : memory.oom_control(oom_kill) 1
> /
> memcg bar : memory.oom_control(oom_kill) 1
>
> For the non-root memcg, its memory.oom_control(oom_kill) includes its
> descendants' oom_kill, but for root memcg, it doesn't include its
> descendants' oom_kill. That means, memory.oom_control(oom_kill) has
> different meanings in different memcgs. That is inconsistent. Then the user
> has to know whether the memcg is root or not.
>
> If we can't fully support it in cgroup1, for example by adding
> memory.events.local into cgroup1 as well, then let's don't touch
> its original behavior.
>
> Fixes: 9852ae3fe529 ("mm, memcg: consider subtrees in memory.events")
> Reviewed-by: Shakeel Butt <shakeelb@google.com>
> Cc: Chris Down <chris@chrisdown.name>
> Cc: Michal Hocko <mhocko@kernel.org>
> Cc: Johannes Weiner <hannes@cmpxchg.org>
> Signed-off-by: Yafang Shao <laoar.shao@gmail.com>
Acked-by: Michal Hocko <mhocko@suse.com>
and sorry to distract you to a cgroup generic solution without doing my
homework and double checking it is possible.
> ---
> include/linux/memcontrol.h | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
> index d275c72c4f8e..977edd3b7bd8 100644
> --- a/include/linux/memcontrol.h
> +++ b/include/linux/memcontrol.h
> @@ -783,6 +783,8 @@ static inline void memcg_memory_event(struct mem_cgroup *memcg,
> atomic_long_inc(&memcg->memory_events[event]);
> cgroup_file_notify(&memcg->events_file);
>
> + if (!cgroup_subsys_on_dfl(memory_cgrp_subsys))
> + break;
> if (cgrp_dfl_root.flags & CGRP_ROOT_MEMORY_LOCAL_EVENTS)
> break;
> } while ((memcg = parent_mem_cgroup(memcg)) &&
> --
> 2.18.2
--
Michal Hocko
SUSE Labs
^ permalink raw reply [flat|nested] 6+ messages in thread* Re: [PATCH resend] mm, memcg: fix inconsistent oom event behavior
2020-05-02 14:10 [PATCH resend] mm, memcg: fix inconsistent oom event behavior Yafang Shao
` (2 preceding siblings ...)
2020-05-04 7:54 ` Michal Hocko
@ 2020-05-04 23:03 ` Andrew Morton
2020-05-05 7:29 ` Michal Hocko
3 siblings, 1 reply; 6+ messages in thread
From: Andrew Morton @ 2020-05-04 23:03 UTC (permalink / raw)
To: Yafang Shao; +Cc: linux-mm, Chris Down, Michal Hocko, Johannes Weiner
On Sat, 2 May 2020 10:10:55 -0400 Yafang Shao <laoar.shao@gmail.com> wrote:
> A recent commit 9852ae3fe529 ("mm, memcg: consider subtrees in
> memory.events") changes the behavior of memcg events, which will
> consider subtrees in memory.events. But oom_kill event is a special one
> as it is used in both cgroup1 and cgroup2. In cgroup1, it is displayed
> in memory.oom_control. The file memory.oom_control is in both root memcg
> and non root memcg, that is different with memory.event as it only in
> non-root memcg. That commit is okay for cgroup2, but it is not okay for
> cgroup1 as it will cause inconsistent behavior between root memcg and
> non-root memcg.
>
> Here's an example on why this behavior is inconsistent in cgroup1.
> root memcg
> /
> memcg foo
> /
> memcg bar
>
> Suppose there's an oom_kill in memcg bar, then the oon_kill will be
>
> root memcg : memory.oom_control(oom_kill) 0
> /
> memcg foo : memory.oom_control(oom_kill) 1
> /
> memcg bar : memory.oom_control(oom_kill) 1
>
> For the non-root memcg, its memory.oom_control(oom_kill) includes its
> descendants' oom_kill, but for root memcg, it doesn't include its
> descendants' oom_kill. That means, memory.oom_control(oom_kill) has
> different meanings in different memcgs. That is inconsistent. Then the user
> has to know whether the memcg is root or not.
>
> If we can't fully support it in cgroup1, for example by adding
> memory.events.local into cgroup1 as well, then let's don't touch
> its original behavior.
>
> Fixes: 9852ae3fe529 ("mm, memcg: consider subtrees in memory.events")
Nearly a year ago. Should we backport this into earlier kernels?
> --- a/include/linux/memcontrol.h
> +++ b/include/linux/memcontrol.h
> @@ -783,6 +783,8 @@ static inline void memcg_memory_event(struct mem_cgroup *memcg,
> atomic_long_inc(&memcg->memory_events[event]);
> cgroup_file_notify(&memcg->events_file);
>
> + if (!cgroup_subsys_on_dfl(memory_cgrp_subsys))
> + break;
> if (cgrp_dfl_root.flags & CGRP_ROOT_MEMORY_LOCAL_EVENTS)
> break;
> } while ((memcg = parent_mem_cgroup(memcg)) &&
^ permalink raw reply [flat|nested] 6+ messages in thread* Re: [PATCH resend] mm, memcg: fix inconsistent oom event behavior
2020-05-04 23:03 ` Andrew Morton
@ 2020-05-05 7:29 ` Michal Hocko
0 siblings, 0 replies; 6+ messages in thread
From: Michal Hocko @ 2020-05-05 7:29 UTC (permalink / raw)
To: Andrew Morton; +Cc: Yafang Shao, linux-mm, Chris Down, Johannes Weiner
On Mon 04-05-20 16:03:45, Andrew Morton wrote:
> On Sat, 2 May 2020 10:10:55 -0400 Yafang Shao <laoar.shao@gmail.com> wrote:
>
> > A recent commit 9852ae3fe529 ("mm, memcg: consider subtrees in
> > memory.events") changes the behavior of memcg events, which will
> > consider subtrees in memory.events. But oom_kill event is a special one
> > as it is used in both cgroup1 and cgroup2. In cgroup1, it is displayed
> > in memory.oom_control. The file memory.oom_control is in both root memcg
> > and non root memcg, that is different with memory.event as it only in
> > non-root memcg. That commit is okay for cgroup2, but it is not okay for
> > cgroup1 as it will cause inconsistent behavior between root memcg and
> > non-root memcg.
> >
> > Here's an example on why this behavior is inconsistent in cgroup1.
> > root memcg
> > /
> > memcg foo
> > /
> > memcg bar
> >
> > Suppose there's an oom_kill in memcg bar, then the oon_kill will be
> >
> > root memcg : memory.oom_control(oom_kill) 0
> > /
> > memcg foo : memory.oom_control(oom_kill) 1
> > /
> > memcg bar : memory.oom_control(oom_kill) 1
> >
> > For the non-root memcg, its memory.oom_control(oom_kill) includes its
> > descendants' oom_kill, but for root memcg, it doesn't include its
> > descendants' oom_kill. That means, memory.oom_control(oom_kill) has
> > different meanings in different memcgs. That is inconsistent. Then the user
> > has to know whether the memcg is root or not.
> >
> > If we can't fully support it in cgroup1, for example by adding
> > memory.events.local into cgroup1 as well, then let's don't touch
> > its original behavior.
> >
> > Fixes: 9852ae3fe529 ("mm, memcg: consider subtrees in memory.events")
>
> Nearly a year ago. Should we backport this into earlier kernels?
It is a trivial change so I do not see problem marking it for stable.
>
> > --- a/include/linux/memcontrol.h
> > +++ b/include/linux/memcontrol.h
> > @@ -783,6 +783,8 @@ static inline void memcg_memory_event(struct mem_cgroup *memcg,
> > atomic_long_inc(&memcg->memory_events[event]);
> > cgroup_file_notify(&memcg->events_file);
> >
> > + if (!cgroup_subsys_on_dfl(memory_cgrp_subsys))
> > + break;
> > if (cgrp_dfl_root.flags & CGRP_ROOT_MEMORY_LOCAL_EVENTS)
> > break;
> > } while ((memcg = parent_mem_cgroup(memcg)) &&
--
Michal Hocko
SUSE Labs
^ permalink raw reply [flat|nested] 6+ messages in thread