[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20170516131316.GA7834@castle>
Date: Tue, 16 May 2017 14:13:16 +0100
From: Roman Gushchin <guro@...com>
To: Michal Hocko <mhocko@...nel.org>
CC: Johannes Weiner <hannes@...xchg.org>, Tejun Heo <tj@...nel.org>,
Li Zefan <lizefan@...wei.com>,
Vladimir Davydov <vdavydov.dev@...il.com>,
<cgroups@...r.kernel.org>, <linux-doc@...r.kernel.org>,
<linux-kernel@...r.kernel.org>, <linux-mm@...ck.org>
Subject: Re: [PATCH] mm: per-cgroup memory reclaim stats
Hi Michal!
On Tue, May 16, 2017 at 11:29:56AM +0200, Michal Hocko wrote:
> On Thu 11-05-17 20:16:23, Roman Gushchin wrote:
> > Track the following reclaim counters for every memory cgroup:
> > PGREFILL, PGSCAN, PGSTEAL, PGACTIVATE, PGDEACTIVATE, PGLAZYFREE and
> > PGLAZYFREED.
>
> yes, those are definitely useful. I have an old patch to add them as
> well but never managed to clean it up and post...
>
> > These values are exposed using the memory.stats interface of cgroup v2.
>
> Is there any reason to not add them to v1?
Not really, I'm just not sure, if it worth it to change v1 interface here.
If you want, I can add them.
> This should be rather trivial after recent changes from Johannes.
If you're about memcg1_events[]/memcg1_event_names[], they can't be reused,
because the pgscan and pgsteal values are both sums of direct and kswapd values:
e.g. events[PGSTEAL_KSWAPD] + events[PGSTEAL_DIRECT].
>
> > The meaning of each value is the same as for global counters,
> > available using /proc/vmstat.
> >
> > Also, for consistency, rename mem_cgroup_count_vm_event() to
> > count_memcg_event_mm().
> >
> > Signed-off-by: Roman Gushchin <guro@...com>
> > Suggested-by: Johannes Weiner <hannes@...xchg.org>
> > Cc: Johannes Weiner <hannes@...xchg.org>
> > Cc: Tejun Heo <tj@...nel.org>
> > Cc: Li Zefan <lizefan@...wei.com>
> > Cc: Michal Hocko <mhocko@...nel.org>
> > Cc: Vladimir Davydov <vdavydov.dev@...il.com>
> > Cc: cgroups@...r.kernel.org
> > Cc: linux-doc@...r.kernel.org
> > Cc: linux-kernel@...r.kernel.org
> > Cc: linux-mm@...ck.org
>
> the patch itself looks good to me. I will have to double check it after
> I am done with what I am doing currently and then will add my Acked-by
Thank you!
Powered by blists - more mailing lists