Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Date:   Tue, 3 Aug 2021 09:11:25 +0200
From:   Michal Hocko <mhocko@suse.com>
To:     Miaohe Lin <linmiaohe@huawei.com>
Cc:     Roman Gushchin <guro@fb.com>, hannes@cmpxchg.org,
        vdavydov.dev@gmail.com, akpm@linux-foundation.org,
        shakeelb@google.com, willy@infradead.org, alexs@kernel.org,
        richard.weiyang@gmail.com, songmuchun@bytedance.com,
        linux-mm@kvack.org, linux-kernel@vger.kernel.org,
        cgroups@vger.kernel.org
Subject: Re: [PATCH 2/5] mm, memcg: narrow the scope of percpu_charge_mutex
Message-ID: <YQjsHTW+46bG9XsV@dhcp22.suse.cz>
References: <20210729125755.16871-1-linmiaohe@huawei.com>
 <20210729125755.16871-3-linmiaohe@huawei.com>
 <YQNsxVPsRSBZcfGG@carbon.lan>
 <YQOhGs3k9rHx3mmT@dhcp22.suse.cz>
 <4a3c23c4-054c-2896-29c5-8cf9a4deee98@huawei.com>
 <YQi6lOT6j2DtOGlT@carbon.dhcp.thefacebook.com>
 <95629d91-6ae8-b445-e7fc-b51c888cad59@huawei.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <95629d91-6ae8-b445-e7fc-b51c888cad59@huawei.com>
Precedence: bulk

On Tue 03-08-21 14:29:13, Miaohe Lin wrote:
[...]
> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> index 616d1a72ece3..6210b1124929 100644
> --- a/mm/memcontrol.c
> +++ b/mm/memcontrol.c
> @@ -2208,11 +2208,11 @@ static void refill_stock(struct mem_cgroup *memcg, unsigned int nr_pages)
>   */
>  static void drain_all_stock(struct mem_cgroup *root_memcg)
>  {
> -       static DEFINE_MUTEX(percpu_charge_mutex);
>         int cpu, curcpu;
> +       static atomic_t drainer = ATOMIC_INIT(0);
> 
>         /* If someone's already draining, avoid adding running more workers. */
> -       if (!mutex_trylock(&percpu_charge_mutex))
> +       if (atomic_cmpxchg(&drainer, 0, 1) != 0)
>                 return;
>         /*
>          * Notify other cpus that system-wide "drain" is running
> @@ -2244,7 +2244,7 @@ static void drain_all_stock(struct mem_cgroup *root_memcg)
>                 }
>         }
>         put_cpu();
> -       mutex_unlock(&percpu_charge_mutex);
> +       atomic_set(&drainer, 0);

atomic_set doesn't imply memory barrier IIRC. Is this safe?

-- 
Michal Hocko
SUSE Labs