Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Fri, 26 Aug 2022 09:38:19 +0200
From:   Michal Hocko <mhocko@suse.com>
To:     Lu Jialin <lujialin4@huawei.com>
Cc:     Zefan Li <lizefan.x@bytedance.com>, Tejun Heo <tj@kernel.org>,
        Johannes Weiner <hannes@cmpxchg.org>,
        Andrew Morton <akpm@linux-foundation.org>,
        Roman Gushchin <roman.gushchin@linux.dev>,
        Shakeel Butt <shakeelb@google.com>,
        Muchun Song <songmuchun@bytedance.com>,
        Xiu Jianfeng <xiujianfeng@huawei.com>, cgroups@vger.kernel.org,
        linux-kernel@vger.kernel.org, linux-mm@kvack.org
Subject: Re: [RFC 2/2] memcg: Adapt cgroup.top into per-memcg
Message-ID: <Ywh4a0NBL3QNMGYS@dhcp22.suse.cz>
References: <20220826011503.103894-1-lujialin4@huawei.com>
 <20220826011503.103894-3-lujialin4@huawei.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20220826011503.103894-3-lujialin4@huawei.com>
Precedence: bulk

On Fri 26-08-22 09:15:03, Lu Jialin wrote:
> cgroup.top is able to show resource usage information for each cgroups.
> Currently only memory usage is monitored ,including usage,anon,file,
> kmem(Bytes), stats for other resource types would be added as well.
> 
> Show case:
> / # mount -t cgroup2 none /sys/fs/cgroup
> / # cd /sys/fs/cgroup/
> /sys/fs/cgroup # echo "+memory" > cgroup.subtree_control
> /sys/fs/cgroup # mkdir test1
> /sys/fs/cgroup # mkdir test2
> /sys/fs/cgroup # mkdir test3
> /sys/fs/cgroup # echo $$ > test2/cgroup.procs
> /sys/fs/cgroup # cd /test
> /test # ./memcg_malloc 512000 &
> /test # ./memcg_malloc 512000 &
> /test # ./memcg_malloc 512000 &
> /test # cd /sys/fs/cgroup
> /sys/fs/cgroup # echo $$ > test1/cgroup.procs
> /sys/fs/cgroup # cd /test
> /test # ./memcg_malloc 512000 &
> /test # cd /sys/fs/cgroup
> /sys/fs/cgroup # echo $$ > test3/cgroup.procs
> /sys/fs/cgroup # cat cgroup.top
> memory top:
> name            usage           anon            file            kernel
> test2           1974272         1671168         0               270336
> test1           700416          569344          0               94208
> test3           196608          86016           0               86016

This information is already present in memory.stat and can be easily
post-processed to generate similar output. Why is that insufficient?
-- 
Michal Hocko
SUSE Labs