Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Subject: Re: pids.current with invalid value for hours [5.0.0 rc3 git]
To:     Tetsuo Handa <penguin-kernel@i-love.sakura.ne.jp>
Cc:     Tejun Heo <tj@kernel.org>, cgroups@vger.kernel.org,
        Aleksa Sarai <asarai@suse.de>, Jay Kamat <jgkamat@fb.com>,
        Roman Gushchin <guro@fb.com>, Michal Hocko <mhocko@suse.com>,
        Johannes Weiner <hannes@cmpxchg.org>,
        linux-kernel@vger.kernel.org,
        Linus Torvalds <torvalds@linux-foundation.org>
References: <df806a77-3327-9db5-8be2-976fde1c84e5@gmail.com>
 <20190117122535.njcbqhlmzozdkncw@mikami>
 <1d36b181-cbaf-6694-1a31-2f7f55d15675@gmail.com>
 <96ef6615-a5df-30af-b4dc-417a18ca63f1@gmail.com>
 <1cdbef13-564d-61a6-95f4-579d2cad243d@gmail.com>
 <20190125163731.GJ50184@devbig004.ftw2.facebook.com>
 <a95d004a-4358-7efc-6d21-12aac4411b32@gmail.com>
 <480296c4-ed7a-3265-e84a-298e42a0f1d5@I-love.SAKURA.ne.jp>
 <6da6ca69-5a6e-a9f6-d091-f89a8488982a@gmail.com>
 <72aa8863-a534-b8df-6b9e-f69cf4dd5c4d@i-love.sakura.ne.jp>
 <33a07810-6dbc-36be-5bb6-a279773ccf69@i-love.sakura.ne.jp>
 <34e97b46-0792-cc66-e0f2-d72576cdec59@i-love.sakura.ne.jp>
From:   =?UTF-8?Q?Arkadiusz_Mi=c5=9bkiewicz?= <a.miskiewicz@gmail.com>
Message-ID: <2b0c7d6c-c58a-da7d-6f0a-4900694ec2d3@gmail.com>
Date:   Sat, 26 Jan 2019 12:29:29 +0100
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.14; rv:60.0)
 Gecko/20100101 Thunderbird/60.4.0
MIME-Version: 1.0
In-Reply-To: <34e97b46-0792-cc66-e0f2-d72576cdec59@i-love.sakura.ne.jp>
Content-Type: text/plain; charset=utf-8
Content-Language: pl
Content-Transfer-Encoding: 8bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On 26/01/2019 12:09, Tetsuo Handa wrote:
> Arkadiusz, will you try this patch?


Works. Several tries and always getting 0 pids.current after ~1s.

Please use

Reported-by: Arkadiusz Miśkiewicz <arekm@maven.pl>

(using gmail for transport only since vger postmasters aren't reasonable)

And also
Tested-by: Arkadiusz Miśkiewicz <arekm@maven.pl>

Testing was done on:
5.0.0-rc3-00104-gc04e2a780caf-dirty

Thanks!

> 
> From 48744b6339cf649a69b55997e138c17df1ecc897 Mon Sep 17 00:00:00 2001
> From: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
> Date: Sat, 26 Jan 2019 20:00:51 +0900
> Subject: [PATCH] oom, oom_reaper: do not enqueue same task twice
> 
> Arkadiusz reported that enabling memcg's group oom killing causes
> strange memcg statistics where there is no task in a memcg despite
> the number of tasks in that memcg is not 0. It turned out that there
> is a bug in wake_oom_reaper() which allows enqueuing same task twice
> which makes impossible to decrease the number of tasks in that memcg
> due to a refcount leak.
> 
> This bug existed since the OOM reaper became invokable from
> task_will_free_mem(current) path in out_of_memory() in Linux 4.7,
> but memcg's group oom killing made it easier to trigger this bug by
> calling wake_oom_reaper() on the same task from one out_of_memory()
> request.
> 
> Fix this bug using an approach used by commit 855b018325737f76
> ("oom, oom_reaper: disable oom_reaper for oom_kill_allocating_task").
> Since task_will_free_mem(p) == false if p->mm == NULL, we can assume that
> p->mm != NULL when wake_oom_reaper() is called from task_will_free_mem()
> paths. As a side effect of this patch, this patch also avoids enqueuing
> multiple threads sharing memory via task_will_free_mem(current) path.
> 
> Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
> Reported-by: Arkadiusz Miśkiewicz <a.miskiewicz@gmail.com>
> Fixes: af8e15cc85a25315 ("oom, oom_reaper: do not enqueue task if it is on the oom_reaper_list head")
> ---
>  mm/oom_kill.c | 28 +++++++++++++---------------
>  1 file changed, 13 insertions(+), 15 deletions(-)
> 
> diff --git a/mm/oom_kill.c b/mm/oom_kill.c
> index f0e8cd9..457f240 100644
> --- a/mm/oom_kill.c
> +++ b/mm/oom_kill.c
> @@ -505,14 +505,6 @@ bool __oom_reap_task_mm(struct mm_struct *mm)
>  	struct vm_area_struct *vma;
>  	bool ret = true;
>  
> -	/*
> -	 * Tell all users of get_user/copy_from_user etc... that the content
> -	 * is no longer stable. No barriers really needed because unmapping
> -	 * should imply barriers already and the reader would hit a page fault
> -	 * if it stumbled over a reaped memory.
> -	 */
> -	set_bit(MMF_UNSTABLE, &mm->flags);
> -
>  	for (vma = mm->mmap ; vma; vma = vma->vm_next) {
>  		if (!can_madv_dontneed_vma(vma))
>  			continue;
> @@ -645,10 +637,15 @@ static int oom_reaper(void *unused)
>  	return 0;
>  }
>  
> -static void wake_oom_reaper(struct task_struct *tsk)
> +static void wake_oom_reaper(struct task_struct *tsk, struct mm_struct *mm)
>  {
> -	/* tsk is already queued? */
> -	if (tsk == oom_reaper_list || tsk->oom_reaper_list)
> +	/*
> +	 * Tell all users of get_user/copy_from_user etc... that the content
> +	 * is no longer stable. No barriers really needed because unmapping
> +	 * should imply barriers already and the reader would hit a page fault
> +	 * if it stumbled over a reaped memory.
> +	 */
> +	if (test_and_set_bit(MMF_UNSTABLE, &mm->flags))
>  		return;
>  
>  	get_task_struct(tsk);
> @@ -668,7 +665,8 @@ static int __init oom_init(void)
>  }
>  subsys_initcall(oom_init)
>  #else
> -static inline void wake_oom_reaper(struct task_struct *tsk)
> +static inline void wake_oom_reaper(struct task_struct *tsk,
> +				   struct mm_struct *mm)
>  {
>  }
>  #endif /* CONFIG_MMU */
> @@ -915,7 +913,7 @@ static void __oom_kill_process(struct task_struct *victim)
>  	rcu_read_unlock();
>  
>  	if (can_oom_reap)
> -		wake_oom_reaper(victim);
> +		wake_oom_reaper(victim, mm);
>  
>  	mmdrop(mm);
>  	put_task_struct(victim);
> @@ -955,7 +953,7 @@ static void oom_kill_process(struct oom_control *oc, const char *message)
>  	task_lock(p);
>  	if (task_will_free_mem(p)) {
>  		mark_oom_victim(p);
> -		wake_oom_reaper(p);
> +		wake_oom_reaper(p, p->mm);
>  		task_unlock(p);
>  		put_task_struct(p);
>  		return;
> @@ -1085,7 +1083,7 @@ bool out_of_memory(struct oom_control *oc)
>  	 */
>  	if (task_will_free_mem(current)) {
>  		mark_oom_victim(current);
> -		wake_oom_reaper(current);
> +		wake_oom_reaper(current, current->mm);
>  		return true;
>  	}
>  
> 


-- 
Arkadiusz Miśkiewicz, arekm / ( maven.pl | pld-linux.org )