linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Minchan Kim <minchan@kernel.org>
To: David Rientjes <rientjes@google.com>
Cc: Luigi Semenzato <semenzato@google.com>,
	linux-mm@kvack.org, Dan Magenheimer <dan.magenheimer@oracle.com>,
	KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>,
	Sonny Rao <sonnyrao@google.com>
Subject: Re: zram OOM behavior
Date: Thu, 1 Nov 2012 11:11:45 +0900	[thread overview]
Message-ID: <20121101021145.GF26256@bbox> (raw)
In-Reply-To: <alpine.DEB.2.00.1210311151341.8809@chino.kir.corp.google.com>

On Wed, Oct 31, 2012 at 11:54:07AM -0700, David Rientjes wrote:
> On Wed, 31 Oct 2012, Minchan Kim wrote:
> 
> > It sounds right in your kernel but principal problem is min_filelist_kbytes patch.
> > If normal exited process in exit path requires a page and there is no free page
> > any more, it ends up going to OOM path after try to reclaim memory several time.
> > Then,
> > In select_bad_process,
> > 
> >         if (task->flags & PF_EXITING) {
> >                if (task == current)             <== true
> >                         return OOM_SCAN_SELECT;
> > In oom_kill_process,
> > 
> >         if (p->flags & PF_EXITING)
> >                 set_tsk_thread_flag(p, TIF_MEMDIE);
> > 
> > At last, normal exited process would get a free page.
> > 
> 
> select_bad_process() won't actually select the process for oom kill, 
> though, if there are other PF_EXITING threads other than current.  So if 
> multiple threads are page faulting on tsk->robust_list, then no thread 
> ends up getting killed.  The temporary workaround would be to do a kill -9 

If mutiple threads are page faulting and try to allocate memory, then they
should go to oom path and they will reach following code.

        if (task->flags & PF_EXITING) {
               if (task == current)
                        return OOM_SCAN_SELECT;

So, the thread can access reseved memory pool and page fault will succeed.

> so that the logic in out_of_memory() could immediately give such threads 
> access to memory reserves so the page fault will succeed.  The real fix 
> would be to audit all possible cases in between setting 
> tsk->flags |= PF_EXITING and tsk->mm = NULL that could cause a memory 
> allocation and make exemptions for them in oom_scan_process_thread().
> 
> --
> To unsubscribe, send a message with 'unsubscribe linux-mm' in
> the body to majordomo@kvack.org.  For more info on Linux MM,
> see: http://www.linux-mm.org/ .
> Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

-- 
Kind regards,
Minchan Kim

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

  parent reply	other threads:[~2012-11-01  2:05 UTC|newest]

Thread overview: 67+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-09-28 17:32 zram OOM behavior Luigi Semenzato
2012-10-03 13:30 ` Konrad Rzeszutek Wilk
     [not found]   ` <CAA25o9SwO209DD6CUx-LzhMt9XU6niGJ-fBPmgwfcrUvf0BPWA@mail.gmail.com>
2012-10-12 23:30     ` Luigi Semenzato
2012-10-15 14:44 ` Minchan Kim
2012-10-15 18:54   ` Luigi Semenzato
2012-10-16  6:18     ` Minchan Kim
2012-10-16 17:36       ` Luigi Semenzato
2012-10-19 17:49         ` Luigi Semenzato
2012-10-22 23:53           ` Minchan Kim
2012-10-23  0:40             ` Luigi Semenzato
2012-10-23  6:03             ` David Rientjes
2012-10-29 18:26               ` Luigi Semenzato
2012-10-29 19:00                 ` David Rientjes
2012-10-29 22:36                   ` Luigi Semenzato
2012-10-29 22:52                     ` David Rientjes
2012-10-29 23:23                       ` Luigi Semenzato
2012-10-29 23:34                         ` Luigi Semenzato
2012-10-30  0:18                     ` Minchan Kim
2012-10-30  0:45                       ` Luigi Semenzato
2012-10-30  5:41                         ` David Rientjes
2012-10-30 19:12                           ` Luigi Semenzato
2012-10-30 20:30                             ` Luigi Semenzato
2012-10-30 22:32                               ` Luigi Semenzato
2012-10-31 18:42                                 ` David Rientjes
2012-10-30 22:37                               ` Sonny Rao
2012-10-31  4:46                               ` David Rientjes
2012-10-31  6:14                                 ` Luigi Semenzato
2012-10-31  6:28                                   ` Luigi Semenzato
2012-10-31 18:45                                     ` David Rientjes
2012-10-31  0:57                             ` Minchan Kim
2012-10-31  1:06                               ` Luigi Semenzato
2012-10-31  1:27                                 ` Minchan Kim
2012-10-31  3:49                                   ` Luigi Semenzato
2012-10-31  7:24                                     ` Minchan Kim
2012-10-31 16:07                                       ` Luigi Semenzato
2012-10-31 17:49                                         ` Mandeep Singh Baines
2012-10-31 18:54                               ` David Rientjes
2012-10-31 21:40                                 ` Luigi Semenzato
2012-11-01  2:11                                 ` Minchan Kim [this message]
2012-11-01  4:38                                   ` David Rientjes
2012-11-01  5:18                                     ` Minchan Kim
2012-11-01  2:43                                 ` Minchan Kim
2012-11-01  4:48                                   ` David Rientjes
2012-11-01  5:26                                     ` Minchan Kim
2012-11-01  8:28                                     ` Mel Gorman
2012-11-01 15:57                                       ` Luigi Semenzato
2012-11-01 15:58                                         ` Luigi Semenzato
2012-11-01 21:48                                           ` David Rientjes
2012-11-01 17:50                                     ` Luigi Semenzato
2012-11-01 21:50                                       ` David Rientjes
2012-11-01 21:58                                         ` [patch] mm, oom: allow exiting threads to have access to memory reserves David Rientjes
2012-11-01 22:43                                           ` Andrew Morton
2012-11-01 23:05                                             ` David Rientjes
2012-11-01 23:06                                             ` Luigi Semenzato
2012-11-01 22:04                                         ` zram OOM behavior Luigi Semenzato
2012-11-01 22:25                                           ` David Rientjes
  -- strict thread matches above, loose matches on Subject: below --
2012-11-02  6:39 Minchan Kim
2012-11-02  8:30 ` Mel Gorman
2012-11-02 22:36   ` Minchan Kim
2012-11-05 14:46     ` Mel Gorman
2012-11-06  0:25       ` Minchan Kim
2012-11-06  8:58         ` Mel Gorman
2012-11-06 10:17           ` Minchan Kim
2012-11-09  9:50             ` Mel Gorman
2012-11-12 13:32               ` Minchan Kim
2012-11-12 14:06                 ` Mel Gorman
2012-11-13 13:31                   ` Minchan Kim

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=20121101021145.GF26256@bbox \
    --to=minchan@kernel.org \
    --cc=dan.magenheimer@oracle.com \
    --cc=kosaki.motohiro@jp.fujitsu.com \
    --cc=linux-mm@kvack.org \
    --cc=rientjes@google.com \
    --cc=semenzato@google.com \
    --cc=sonnyrao@google.com \
    /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 NNTP newsgroup(s).