From: Michal Hocko <mhocko@kernel.org>
To: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Cc: akpm@linux-foundation.org, rientjes@google.com, mgorman@suse.de,
oleg@redhat.com, torvalds@linux-foundation.org, hughd@google.com,
andrea@kernel.org, riel@redhat.com, linux-mm@kvack.org,
linux-kernel@vger.kernel.org
Subject: Re: [PATCH 3/5] oom: clear TIF_MEMDIE after oom_reaper managed to unmap the address space
Date: Thu, 4 Feb 2016 17:31:13 +0100 [thread overview]
Message-ID: <20160204163113.GF14425@dhcp22.suse.cz> (raw)
In-Reply-To: <201602050008.HEG12919.FFOMOHVtQFSLJO@I-love.SAKURA.ne.jp>
On Fri 05-02-16 00:08:25, Tetsuo Handa wrote:
> Michal Hocko wrote:
> > > > + /*
> > > > + * Clear TIF_MEMDIE because the task shouldn't be sitting on a
> > > > + * reasonably reclaimable memory anymore. OOM killer can continue
> > > > + * by selecting other victim if unmapping hasn't led to any
> > > > + * improvements. This also means that selecting this task doesn't
> > > > + * make any sense.
> > > > + */
> > > > + tsk->signal->oom_score_adj = OOM_SCORE_ADJ_MIN;
> > > > + exit_oom_victim(tsk);
> > >
> > > I noticed that updating only one thread group's oom_score_adj disables
> > > further wake_oom_reaper() calls due to rough-grained can_oom_reap check at
> > >
> > > p->signal->oom_score_adj == OOM_SCORE_ADJ_MIN
> > >
> > > in oom_kill_process(). I think we need to either update all thread groups'
> > > oom_score_adj using the reaped mm equally or use more fine-grained can_oom_reap
> > > check which ignores OOM_SCORE_ADJ_MIN if all threads in that thread group are
> > > dying or exiting.
> >
> > I do not understand. Why would you want to reap the mm again when
> > this has been done already? The mm is shared, right?
>
> The mm is shared between previous victim and next victim, but these victims
> are in different thread groups. The OOM killer selects next victim whose mm
> was already reaped due to sharing previous victim's memory.
OK, now I got your point. From your previous email it sounded like you
were talking about oom_reaper and its invocation which is was confusing.
> We don't want the OOM killer to select such next victim.
Yes, selecting such a task doesn't make much sense. It has been killed
so it has fatal_signal_pending. If it wanted to allocate it would get
TIF_MEMDIE already and it's address space has been reaped so there is
nothing to free left. These CLONE_VM without CLONE_SIGHAND is really
crazy combo, it is just causing troubles all over and I am not convinced
it is actually that helpful </rant>.
> Maybe set MMF_OOM_REAP_DONE on
> the previous victim's mm and check it instead of TIF_MEMDIE when selecting
> a victim? That will also avoid problems caused by clearing TIF_MEMDIE?
Hmm, it doesn't seem we are under MMF_ availabel bits pressure right now
so using the flag sounds like the easiest way to go. Then we even do not
have to play with OOM_SCORE_ADJ_MIN which might be updated from the
userspace after the oom reaper has done that. Care to send a patch?
Thanks!
--
Michal Hocko
SUSE Labs
next prev parent reply other threads:[~2016-02-04 16:31 UTC|newest]
Thread overview: 48+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-02-03 13:13 [PATCH 0/5] oom reaper v5 Michal Hocko
2016-02-03 13:13 ` [PATCH 1/5] mm, oom: introduce oom reaper Michal Hocko
2016-02-03 23:48 ` David Rientjes
2016-02-04 6:41 ` Michal Hocko
2016-02-06 13:22 ` Tetsuo Handa
2016-02-15 20:50 ` Michal Hocko
2016-02-03 13:13 ` [PATCH 2/5] oom reaper: handle mlocked pages Michal Hocko
2016-02-03 23:57 ` David Rientjes
2016-02-23 1:36 ` David Rientjes
2016-02-23 13:21 ` Michal Hocko
2016-02-29 3:19 ` Hugh Dickins
2016-02-29 13:41 ` Michal Hocko
2016-03-08 13:40 ` Michal Hocko
2016-03-08 20:07 ` Hugh Dickins
2016-03-09 8:26 ` Michal Hocko
2016-02-03 13:13 ` [PATCH 3/5] oom: clear TIF_MEMDIE after oom_reaper managed to unmap the address space Michal Hocko
2016-02-04 14:22 ` Tetsuo Handa
2016-02-04 14:43 ` Michal Hocko
2016-02-04 15:08 ` Tetsuo Handa
2016-02-04 16:31 ` Michal Hocko [this message]
2016-02-05 11:14 ` Tetsuo Handa
2016-02-06 8:30 ` Michal Hocko
2016-02-06 11:23 ` Tetsuo Handa
2016-02-15 20:47 ` Michal Hocko
2016-02-06 6:45 ` Michal Hocko
2016-02-06 14:33 ` Tetsuo Handa
2016-02-15 20:40 ` [PATCH 3.1/5] oom: make oom_reaper freezable Michal Hocko
2016-02-03 13:13 ` [PATCH 4/5] mm, oom_reaper: report success/failure Michal Hocko
2016-02-03 23:10 ` David Rientjes
2016-02-04 6:46 ` Michal Hocko
2016-02-04 22:31 ` David Rientjes
2016-02-05 9:26 ` Michal Hocko
2016-02-06 6:34 ` Michal Hocko
2016-02-03 13:14 ` [PATCH 5/5] mm, oom_reaper: implement OOM victims queuing Michal Hocko
2016-02-04 10:49 ` Tetsuo Handa
2016-02-04 14:53 ` Michal Hocko
2016-02-06 5:54 ` Tetsuo Handa
2016-02-06 8:37 ` Michal Hocko
2016-02-06 15:33 ` Tetsuo Handa
2016-02-15 20:15 ` Michal Hocko
2016-02-16 11:11 ` Tetsuo Handa
2016-02-16 15:53 ` Michal Hocko
2016-02-17 9:48 ` [PATCH 6/5] oom, oom_reaper: disable oom_reaper for Michal Hocko
2016-02-17 10:41 ` Tetsuo Handa
2016-02-17 11:33 ` Michal Hocko
2016-02-19 18:34 ` Michal Hocko
2016-02-20 2:32 ` [PATCH 6/5] oom, oom_reaper: disable oom_reaper for oom_kill_allocating_task Tetsuo Handa
2016-02-22 9:41 ` Michal Hocko
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20160204163113.GF14425@dhcp22.suse.cz \
--to=mhocko@kernel.org \
--cc=akpm@linux-foundation.org \
--cc=andrea@kernel.org \
--cc=hughd@google.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mgorman@suse.de \
--cc=oleg@redhat.com \
--cc=penguin-kernel@I-love.SAKURA.ne.jp \
--cc=riel@redhat.com \
--cc=rientjes@google.com \
--cc=torvalds@linux-foundation.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).