From: Andrew Morton <akpm@linux-foundation.org>
To: Changli Gao <xiaosuo@gmail.com>
Cc: "Eric Dumazet" <eric.dumazet@gmail.com>,
"Américo Wang" <xiyou.wangcong@gmail.com>,
"Jiri Slaby" <jslaby@suse.cz>, azurIt <azurit@pobox.sk>,
linux-kernel@vger.kernel.org, linux-mm@kvack.org,
linux-fsdevel@vger.kernel.org, "Jiri Slaby" <jirislaby@gmail.com>
Subject: Re: Regression from 2.6.36
Date: Tue, 12 Apr 2011 18:31:32 -0700 [thread overview]
Message-ID: <20110412183132.a854bffc.akpm@linux-foundation.org> (raw)
In-Reply-To: <BANLkTincoaxp5Soe6O-eb8LWpgra=k2NsQ@mail.gmail.com>
On Wed, 13 Apr 2011 09:23:11 +0800 Changli Gao <xiaosuo@gmail.com> wrote:
> On Wed, Apr 13, 2011 at 6:49 AM, Andrew Morton
> <akpm@linux-foundation.org> wrote:
> >
> > It's somewhat unclear (to me) what caused this regression.
> >
> > Is it because the kernel is now doing large kmalloc()s for the fdtable,
> > and this makes the page allocator go nuts trying to satisfy high-order
> > page allocation requests?
> >
> > Is it because the kernel now will usually free the fdtable
> > synchronously within the rcu callback, rather than deferring this to a
> > workqueue?
> >
> > The latter seems unlikely, so I'm thinking this was a case of
> > high-order-allocations-considered-harmful?
> >
>
> Maybe, but I am not sure. Maybe my patch causes too many inner
> fragments. For example, when asking for 5 pages, get 8 pages, and 3
> pages are wasted, then memory thrash happens finally.
That theory sounds less likely, but could be tested by using
alloc_pages_exact().
WARNING: multiple messages have this Message-ID (diff)
From: Andrew Morton <akpm@linux-foundation.org>
To: Changli Gao <xiaosuo@gmail.com>
Cc: "Eric Dumazet" <eric.dumazet@gmail.com>,
"Américo Wang" <xiyou.wangcong@gmail.com>,
"Jiri Slaby" <jslaby@suse.cz>, azurIt <azurit@pobox.sk>,
linux-kernel@vger.kernel.org, linux-mm@kvack.org,
linux-fsdevel@vger.kernel.org, "Jiri Slaby" <jirislaby@gmail.com>
Subject: Re: Regression from 2.6.36
Date: Tue, 12 Apr 2011 18:31:32 -0700 [thread overview]
Message-ID: <20110412183132.a854bffc.akpm@linux-foundation.org> (raw)
In-Reply-To: <BANLkTincoaxp5Soe6O-eb8LWpgra=k2NsQ@mail.gmail.com>
On Wed, 13 Apr 2011 09:23:11 +0800 Changli Gao <xiaosuo@gmail.com> wrote:
> On Wed, Apr 13, 2011 at 6:49 AM, Andrew Morton
> <akpm@linux-foundation.org> wrote:
> >
> > It's somewhat unclear (to me) what caused this regression.
> >
> > Is it because the kernel is now doing large kmalloc()s for the fdtable,
> > and this makes the page allocator go nuts trying to satisfy high-order
> > page allocation requests?
> >
> > Is it because the kernel now will usually free the fdtable
> > synchronously within the rcu callback, rather than deferring this to a
> > workqueue?
> >
> > The latter seems unlikely, so I'm thinking this was a case of
> > high-order-allocations-considered-harmful?
> >
>
> Maybe, but I am not sure. Maybe my patch causes too many inner
> fragments. For example, when asking for 5 pages, get 8 pages, and 3
> pages are wasted, then memory thrash happens finally.
That theory sounds less likely, but could be tested by using
alloc_pages_exact().
--
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:[~2011-04-13 1:29 UTC|newest]
Thread overview: 98+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-03-15 13:25 Regression from 2.6.36 azurIt
2011-03-17 0:15 ` Greg KH
2011-03-17 0:53 ` Dave Jones
2011-03-17 13:30 ` azurIt
2011-04-07 10:01 ` azurIt
2011-04-07 10:19 ` Jiri Slaby
2011-04-07 10:19 ` Jiri Slaby
2011-04-07 10:19 ` Jiri Slaby
2011-04-07 11:21 ` Américo Wang
2011-04-07 11:21 ` Américo Wang
2011-04-07 11:57 ` Eric Dumazet
2011-04-07 11:57 ` Eric Dumazet
2011-04-07 11:57 ` Eric Dumazet
2011-04-07 12:13 ` Eric Dumazet
2011-04-07 12:13 ` Eric Dumazet
2011-04-07 12:13 ` Eric Dumazet
2011-04-07 15:27 ` Changli Gao
2011-04-07 15:36 ` Eric Dumazet
2011-04-07 15:36 ` Eric Dumazet
2011-04-07 15:36 ` Eric Dumazet
2011-04-12 22:49 ` Andrew Morton
2011-04-12 22:49 ` Andrew Morton
2011-04-13 1:23 ` Changli Gao
2011-04-13 1:23 ` Changli Gao
2011-04-13 1:31 ` Andrew Morton [this message]
2011-04-13 1:31 ` Andrew Morton
2011-04-13 2:37 ` Eric Dumazet
2011-04-13 2:37 ` Eric Dumazet
2011-04-13 2:37 ` Eric Dumazet
2011-04-13 6:54 ` Regarding memory fragmentation using malloc Pintu Agarwal
2011-04-13 6:54 ` Pintu Agarwal
2011-04-13 11:44 ` Américo Wang
2011-04-13 11:44 ` Américo Wang
2011-04-13 13:56 ` Pintu Agarwal
2011-04-13 13:56 ` Pintu Agarwal
2011-04-13 15:25 ` Michal Nazarewicz
2011-04-13 15:25 ` Michal Nazarewicz
2011-04-14 6:44 ` Pintu Agarwal
2011-04-14 6:44 ` Pintu Agarwal
2011-04-14 10:47 ` Michal Nazarewicz
2011-04-14 10:47 ` Michal Nazarewicz
2011-04-14 12:24 ` Pintu Agarwal
2011-04-14 12:24 ` Pintu Agarwal
2011-04-14 12:31 ` Michal Nazarewicz
2011-04-14 12:31 ` Michal Nazarewicz
2011-04-13 21:16 ` Regression from 2.6.36 Andrew Morton
2011-04-13 21:16 ` Andrew Morton
2011-04-13 21:24 ` Andrew Morton
2011-04-13 21:24 ` Andrew Morton
2011-04-13 21:24 ` Andrew Morton
2011-04-19 19:29 ` azurIt
2011-04-19 19:29 ` azurIt
2011-04-19 19:55 ` Andrew Morton
2011-04-19 19:55 ` Andrew Morton
2011-04-13 21:44 ` David Rientjes
2011-04-13 21:44 ` David Rientjes
2011-04-13 21:54 ` Andrew Morton
2011-04-13 21:54 ` Andrew Morton
2011-04-14 2:10 ` Eric Dumazet
2011-04-14 2:10 ` Eric Dumazet
2011-04-14 2:10 ` Eric Dumazet
2011-04-14 5:28 ` Andrew Morton
2011-04-14 5:28 ` Andrew Morton
2011-04-14 6:31 ` Eric Dumazet
2011-04-14 6:31 ` Eric Dumazet
2011-04-14 6:31 ` Eric Dumazet
2011-04-14 9:08 ` azurIt
2011-04-14 9:08 ` azurIt
2011-04-14 10:27 ` Eric Dumazet
2011-04-14 10:27 ` Eric Dumazet
2011-04-14 10:27 ` Eric Dumazet
2011-04-14 10:31 ` azurIt
2011-04-14 10:31 ` azurIt
2011-04-14 10:31 ` azurIt
2011-04-14 10:25 ` Mel Gorman
2011-04-15 9:59 ` azurIt
2011-04-15 9:59 ` azurIt
2011-04-15 9:59 ` azurIt
2011-04-15 10:47 ` Mel Gorman
2011-04-15 10:47 ` Mel Gorman
2011-04-15 10:56 ` azurIt
2011-04-15 10:56 ` azurIt
2011-04-15 10:56 ` azurIt
2011-04-15 11:17 ` Mel Gorman
2011-04-15 11:17 ` Mel Gorman
2011-04-15 11:36 ` azurIt
2011-04-15 11:36 ` azurIt
2011-04-15 11:36 ` azurIt
2011-04-15 13:01 ` Mel Gorman
2011-04-15 13:01 ` Mel Gorman
2011-04-15 13:21 ` azurIt
2011-04-15 13:21 ` azurIt
2011-04-15 13:21 ` azurIt
2011-04-15 14:15 ` Mel Gorman
2011-04-15 14:15 ` Mel Gorman
2011-04-08 12:25 ` azurIt
2011-04-08 12:25 ` azurIt
2011-04-08 12:25 ` azurIt
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=20110412183132.a854bffc.akpm@linux-foundation.org \
--to=akpm@linux-foundation.org \
--cc=azurit@pobox.sk \
--cc=eric.dumazet@gmail.com \
--cc=jirislaby@gmail.com \
--cc=jslaby@suse.cz \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=xiaosuo@gmail.com \
--cc=xiyou.wangcong@gmail.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.