From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
To: Minchan Kim <minchan.kim@gmail.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>,
Andrew Lutomirski <luto@mit.edu>,
KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>,
fengguang.wu@intel.com, andi@firstfloor.org, linux-mm@kvack.org,
linux-kernel@vger.kernel.org, mgorman@suse.de,
hannes@cmpxchg.org, riel@redhat.com
Subject: Re: Kernel falls apart under light memory pressure (i.e. linking vmlinux)
Date: Fri, 20 May 2011 16:43:54 +0900 [thread overview]
Message-ID: <20110520164354.d43be406.kamezawa.hiroyu@jp.fujitsu.com> (raw)
In-Reply-To: <BANLkTinJbYrQoye7qjPzPxP8_deCSK0g7w@mail.gmail.com>
On Fri, 20 May 2011 14:36:13 +0900
Minchan Kim <minchan.kim@gmail.com> wrote:
> On Fri, May 20, 2011 at 2:08 PM, KAMEZAWA Hiroyuki
> <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> > On Fri, 20 May 2011 13:20:15 +0900
> > Minchan Kim <minchan.kim@gmail.com> wrote:
> >
> >> So I want to resolve your problem asap.
> >> We don't have see report about that. Could you do git-bisect?
> >> FYI, Recently, big change of mm is compaction,transparent huge pages.
> >> Kame, could you point out thing related to memcg if you have a mind?
> >>
> >
> > I don't doubt memcg at this stage because it never modify page->flags.
> > Consdering the case, PageActive() is set against off-LRU pages after
> > clear_active_flags() clears it.
> >
> > Hmm, I think I don't understand the lock system fully but...how do you
> > think this ?
> >
> > ==
> >
> > At splitting a hugepage, the routine marks all pmd as "splitting".
> >
> > But assume a racy case where 2 threads run into spit at the
> > same time, one thread wins compound_lock() and do split, another
> > thread should not touch splitted pages.
>
> Sorry. Now I don't have a time to review in detail.
> When I look it roughly, page_lock_anon_vma have to prevent it.
> But Andrea needs current this problem and he will catch something we lost. :)
>
Hmm, maybe I miss something...need to build a test environ on my side.
But I'm not sure I can reproduce it..
Thanks,
-Kame
WARNING: multiple messages have this Message-ID (diff)
From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
To: Minchan Kim <minchan.kim@gmail.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>,
Andrew Lutomirski <luto@mit.edu>,
KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>,
fengguang.wu@intel.com, andi@firstfloor.org, linux-mm@kvack.org,
linux-kernel@vger.kernel.org, mgorman@suse.de,
hannes@cmpxchg.org, riel@redhat.com
Subject: Re: Kernel falls apart under light memory pressure (i.e. linking vmlinux)
Date: Fri, 20 May 2011 16:43:54 +0900 [thread overview]
Message-ID: <20110520164354.d43be406.kamezawa.hiroyu@jp.fujitsu.com> (raw)
In-Reply-To: <BANLkTinJbYrQoye7qjPzPxP8_deCSK0g7w@mail.gmail.com>
On Fri, 20 May 2011 14:36:13 +0900
Minchan Kim <minchan.kim@gmail.com> wrote:
> On Fri, May 20, 2011 at 2:08 PM, KAMEZAWA Hiroyuki
> <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> > On Fri, 20 May 2011 13:20:15 +0900
> > Minchan Kim <minchan.kim@gmail.com> wrote:
> >
> >> So I want to resolve your problem asap.
> >> We don't have see report about that. Could you do git-bisect?
> >> FYI, Recently, big change of mm is compaction,transparent huge pages.
> >> Kame, could you point out thing related to memcg if you have a mind?
> >>
> >
> > I don't doubt memcg at this stage because it never modify page->flags.
> > Consdering the case, PageActive() is set against off-LRU pages after
> > clear_active_flags() clears it.
> >
> > Hmm, I think I don't understand the lock system fully but...how do you
> > think this ?
> >
> > ==
> >
> > At splitting a hugepage, the routine marks all pmd as "splitting".
> >
> > But assume a racy case where 2 threads run into spit at the
> > same time, one thread wins compound_lock() and do split, another
> > thread should not touch splitted pages.
>
> Sorry. Now I don't have a time to review in detail.
> When I look it roughly, page_lock_anon_vma have to prevent it.
> But Andrea needs current this problem and he will catch something we lost. :)
>
Hmm, maybe I miss something...need to build a test environ on my side.
But I'm not sure I can reproduce it..
Thanks,
-Kame
--
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/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
next prev parent reply other threads:[~2011-05-20 7:50 UTC|newest]
Thread overview: 102+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-05-11 22:42 Kernel falls apart under light memory pressure (i.e. linking vmlinux) Andrew Lutomirski
2011-05-11 23:07 ` Andi Kleen
2011-05-11 23:28 ` Andrew Lutomirski
2011-05-12 5:46 ` Andi Kleen
2011-05-12 11:54 ` Andrew Lutomirski
2011-05-14 15:46 ` Andrew Lutomirski
2011-05-14 15:46 ` Andrew Lutomirski
2011-05-14 16:53 ` Andi Kleen
2011-05-14 16:53 ` Andi Kleen
[not found] ` <BANLkTik6SS9NH7XVSRBoCR16_5veY0MKBw@mail.gmail.com>
2011-05-14 17:43 ` Andi Kleen
2011-05-15 1:37 ` Minchan Kim
2011-05-15 15:27 ` Wu Fengguang
2011-05-15 15:27 ` Wu Fengguang
2011-05-15 15:59 ` Andrew Lutomirski
2011-05-15 15:59 ` Andrew Lutomirski
2011-05-15 22:58 ` Minchan Kim
2011-05-15 22:58 ` Minchan Kim
2011-05-16 8:51 ` Mel Gorman
2011-05-16 8:51 ` Mel Gorman
2011-05-15 16:12 ` Andrew Lutomirski
2011-05-15 16:12 ` Andrew Lutomirski
2011-05-17 6:00 ` Wu Fengguang
2011-05-17 6:00 ` Wu Fengguang
2011-05-17 6:35 ` Minchan Kim
2011-05-17 6:35 ` Minchan Kim
2011-05-17 19:22 ` Andrew Lutomirski
2011-05-18 5:17 ` Minchan Kim
2011-05-18 5:17 ` Minchan Kim
2011-05-19 2:15 ` Andrew Lutomirski
2011-05-19 2:30 ` KAMEZAWA Hiroyuki
2011-05-19 2:30 ` KAMEZAWA Hiroyuki
2011-05-19 2:41 ` Andrew Lutomirski
2011-05-19 2:54 ` Minchan Kim
2011-05-19 2:54 ` Minchan Kim
2011-05-19 14:16 ` Andrew Lutomirski
2011-05-20 0:17 ` Minchan Kim
2011-05-20 0:17 ` Minchan Kim
2011-05-20 2:58 ` Andrew Lutomirski
2011-05-20 2:58 ` Andrew Lutomirski
2011-05-20 3:12 ` KOSAKI Motohiro
2011-05-20 3:12 ` KOSAKI Motohiro
2011-05-20 3:38 ` Andrew Lutomirski
2011-05-20 3:38 ` Andrew Lutomirski
2011-05-20 4:20 ` Minchan Kim
2011-05-20 4:20 ` Minchan Kim
2011-05-20 5:08 ` KAMEZAWA Hiroyuki
2011-05-20 5:08 ` KAMEZAWA Hiroyuki
2011-05-20 5:36 ` Minchan Kim
2011-05-20 5:36 ` Minchan Kim
2011-05-20 7:43 ` KAMEZAWA Hiroyuki [this message]
2011-05-20 7:43 ` KAMEZAWA Hiroyuki
2011-05-20 10:11 ` Andrea Arcangeli
2011-05-20 10:11 ` Andrea Arcangeli
2011-05-20 14:11 ` Andrew Lutomirski
2011-05-20 15:33 ` Minchan Kim
2011-05-20 15:33 ` Minchan Kim
2011-05-20 16:01 ` Andrew Lutomirski
2011-05-20 16:01 ` Andrew Lutomirski
2011-05-20 16:19 ` Minchan Kim
2011-05-20 16:19 ` Minchan Kim
2011-05-20 18:09 ` Andrew Lutomirski
2011-05-20 18:40 ` Andrew Lutomirski
2011-05-20 18:40 ` Andrew Lutomirski
2011-05-21 12:04 ` KOSAKI Motohiro
2011-05-21 12:04 ` KOSAKI Motohiro
2011-05-21 13:34 ` Andrew Lutomirski
2011-05-21 13:34 ` Andrew Lutomirski
2011-05-21 14:14 ` KOSAKI Motohiro
2011-05-21 14:14 ` KOSAKI Motohiro
2011-05-21 14:44 ` Minchan Kim
2011-05-21 14:44 ` Minchan Kim
2011-05-22 12:22 ` Andrew Lutomirski
2011-05-22 12:22 ` Andrew Lutomirski
2011-05-22 23:12 ` Minchan Kim
2011-05-22 23:12 ` Minchan Kim
2011-05-23 16:42 ` Andrea Arcangeli
2011-05-23 16:42 ` Andrea Arcangeli
2011-05-23 17:35 ` Mel Gorman
2011-05-23 17:35 ` Mel Gorman
2011-05-24 1:19 ` Andrew Lutomirski
2011-05-24 1:34 ` Minchan Kim
2011-05-24 1:34 ` Minchan Kim
2011-05-24 11:24 ` Andrew Lutomirski
2011-05-24 11:24 ` Andrew Lutomirski
2011-05-24 11:55 ` Andrew Lutomirski
2011-05-25 0:43 ` KOSAKI Motohiro
2011-05-25 0:43 ` KOSAKI Motohiro
2011-05-21 14:31 ` Minchan Kim
2011-05-21 14:31 ` Minchan Kim
2011-05-19 14:51 ` Wu Fengguang
2011-05-19 14:51 ` Wu Fengguang
2011-05-19 15:00 ` Andrew Lutomirski
2011-05-19 15:00 ` Andrew Lutomirski
2011-05-20 0:20 ` Minchan Kim
2011-05-20 0:20 ` Minchan Kim
2011-05-15 22:40 ` Minchan Kim
2011-05-15 22:40 ` Minchan Kim
2011-05-17 5:52 ` Wu Fengguang
2011-05-17 5:52 ` Wu Fengguang
2011-05-17 6:26 ` Minchan Kim
2011-05-17 6:26 ` Minchan Kim
2011-05-20 10:40 ` Andrea Arcangeli
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=20110520164354.d43be406.kamezawa.hiroyu@jp.fujitsu.com \
--to=kamezawa.hiroyu@jp.fujitsu.com \
--cc=aarcange@redhat.com \
--cc=andi@firstfloor.org \
--cc=fengguang.wu@intel.com \
--cc=hannes@cmpxchg.org \
--cc=kosaki.motohiro@jp.fujitsu.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=luto@mit.edu \
--cc=mgorman@suse.de \
--cc=minchan.kim@gmail.com \
--cc=riel@redhat.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.