From: Rik van Riel <riel@redhat.com>
To: Mel Gorman <mgorman@suse.de>
Cc: Thorsten Leemhuis <fedora@leemhuis.info>,
Jiri Slaby <jslaby@suse.cz>,
Valdis.Kletnieks@vt.edu, Jiri Slaby <jirislaby@gmail.com>,
linux-mm@kvack.org, LKML <linux-kernel@vger.kernel.org>,
Andrew Morton <akpm@linux-foundation.org>
Subject: Re: kswapd0: excessive CPU usage
Date: Sun, 04 Nov 2012 11:36:47 -0500 [thread overview]
Message-ID: <5096999F.1040405@redhat.com> (raw)
In-Reply-To: <20121030191843.GH3888@suse.de>
On 10/30/2012 03:18 PM, Mel Gorman wrote:
> restart:
> - wake_all_kswapd(order, zonelist, high_zoneidx,
> + /*
> + * kswapd is woken except when this is a THP request and compaction
> + * is deferred. If we are backing off reclaim/compaction then kswapd
> + * should not be awake aggressively reclaiming with no consumers of
> + * the freed pages
> + */
> + if (!(is_thp_alloc(gfp_mask, order) &&
> + compaction_deferred(preferred_zone, order)))
> + wake_all_kswapd(order, zonelist, high_zoneidx,
> zone_idx(preferred_zone));
What is special about thp allocations here?
Surely other large allocations that keep failing
should get the same treatment, of not waking up
kswapd if compaction is deferred?
--
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>
WARNING: multiple messages have this Message-ID (diff)
From: Rik van Riel <riel@redhat.com>
To: Mel Gorman <mgorman@suse.de>
Cc: Thorsten Leemhuis <fedora@leemhuis.info>,
Jiri Slaby <jslaby@suse.cz>,
Valdis.Kletnieks@vt.edu, Jiri Slaby <jirislaby@gmail.com>,
linux-mm@kvack.org, LKML <linux-kernel@vger.kernel.org>,
Andrew Morton <akpm@linux-foundation.org>
Subject: Re: kswapd0: excessive CPU usage
Date: Sun, 04 Nov 2012 11:36:47 -0500 [thread overview]
Message-ID: <5096999F.1040405@redhat.com> (raw)
In-Reply-To: <20121030191843.GH3888@suse.de>
On 10/30/2012 03:18 PM, Mel Gorman wrote:
> restart:
> - wake_all_kswapd(order, zonelist, high_zoneidx,
> + /*
> + * kswapd is woken except when this is a THP request and compaction
> + * is deferred. If we are backing off reclaim/compaction then kswapd
> + * should not be awake aggressively reclaiming with no consumers of
> + * the freed pages
> + */
> + if (!(is_thp_alloc(gfp_mask, order) &&
> + compaction_deferred(preferred_zone, order)))
> + wake_all_kswapd(order, zonelist, high_zoneidx,
> zone_idx(preferred_zone));
What is special about thp allocations here?
Surely other large allocations that keep failing
should get the same treatment, of not waking up
kswapd if compaction is deferred?
next prev parent reply other threads:[~2012-11-04 16:34 UTC|newest]
Thread overview: 100+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-10-11 8:52 kswapd0: wxcessive CPU usage Jiri Slaby
2012-10-11 8:52 ` Jiri Slaby
2012-10-11 13:44 ` Valdis.Kletnieks
2012-10-11 15:34 ` Jiri Slaby
2012-10-11 15:34 ` Jiri Slaby
2012-10-11 17:56 ` Valdis.Kletnieks
2012-10-11 17:59 ` Jiri Slaby
2012-10-11 17:59 ` Jiri Slaby
2012-10-11 18:19 ` Valdis.Kletnieks
2012-10-11 22:08 ` kswapd0: excessive " Jiri Slaby
2012-10-11 22:08 ` Jiri Slaby
2012-10-12 12:37 ` Jiri Slaby
2012-10-12 12:37 ` Jiri Slaby
2012-10-12 13:57 ` Mel Gorman
2012-10-12 13:57 ` Mel Gorman
2012-10-15 9:54 ` Jiri Slaby
2012-10-15 9:54 ` Jiri Slaby
2012-10-15 11:09 ` Mel Gorman
2012-10-15 11:09 ` Mel Gorman
2012-10-29 10:52 ` Thorsten Leemhuis
2012-10-29 10:52 ` Thorsten Leemhuis
2012-10-30 19:18 ` Mel Gorman
2012-10-30 19:18 ` Mel Gorman
2012-10-31 11:25 ` Thorsten Leemhuis
2012-10-31 11:25 ` Thorsten Leemhuis
2012-10-31 15:04 ` Mel Gorman
2012-10-31 15:04 ` Mel Gorman
2012-11-04 16:36 ` Rik van Riel [this message]
2012-11-04 16:36 ` Rik van Riel
2012-11-02 10:44 ` Zdenek Kabelac
2012-11-02 10:44 ` Zdenek Kabelac
2012-11-02 10:53 ` Jiri Slaby
2012-11-02 10:53 ` Jiri Slaby
2012-11-02 19:45 ` Jiri Slaby
2012-11-02 19:45 ` Jiri Slaby
2012-11-04 11:26 ` Zdenek Kabelac
2012-11-04 11:26 ` Zdenek Kabelac
2012-11-05 14:24 ` [PATCH] Revert "mm: vmscan: scale number of pages reclaimed by reclaim/compaction based on failures" Mel Gorman
2012-11-05 14:24 ` Mel Gorman
2012-11-06 10:15 ` Johannes Hirte
2012-11-06 10:15 ` Johannes Hirte
2012-11-09 8:36 ` Mel Gorman
2012-11-09 8:36 ` Mel Gorman
2012-11-14 21:43 ` Johannes Hirte
2012-11-14 21:43 ` Johannes Hirte
2012-11-09 9:12 ` Mel Gorman
2012-11-09 9:12 ` Mel Gorman
2012-11-09 4:22 ` kswapd0: excessive CPU usage Seth Jennings
2012-11-09 4:22 ` Seth Jennings
2012-11-09 8:07 ` Zdenek Kabelac
2012-11-09 8:07 ` Zdenek Kabelac
2012-11-09 9:06 ` Mel Gorman
2012-11-09 9:06 ` Mel Gorman
2012-11-11 9:13 ` Zdenek Kabelac
2012-11-11 9:13 ` Zdenek Kabelac
2012-11-12 11:37 ` [PATCH] Revert "mm: remove __GFP_NO_KSWAPD" Mel Gorman
2012-11-12 11:37 ` Mel Gorman
2012-11-16 19:14 ` Josh Boyer
2012-11-16 19:14 ` Josh Boyer
2012-11-16 19:51 ` Andrew Morton
2012-11-16 19:51 ` Andrew Morton
2012-11-20 1:43 ` Valdis.Kletnieks
2012-11-16 20:06 ` Mel Gorman
2012-11-16 20:06 ` Mel Gorman
2012-11-20 15:38 ` Josh Boyer
2012-11-20 15:38 ` Josh Boyer
2012-11-20 16:13 ` Bruno Wolff III
2012-11-20 16:13 ` Bruno Wolff III
2012-11-20 17:43 ` Thorsten Leemhuis
2012-11-20 17:43 ` Thorsten Leemhuis
2012-11-23 15:20 ` Thorsten Leemhuis
2012-11-23 15:20 ` Thorsten Leemhuis
2012-11-27 11:12 ` Mel Gorman
2012-11-27 11:12 ` Mel Gorman
2012-11-21 15:08 ` Mel Gorman
2012-11-21 15:08 ` Mel Gorman
2012-11-20 9:18 ` Glauber Costa
2012-11-20 9:18 ` Glauber Costa
2012-11-20 20:18 ` Andrew Morton
2012-11-20 20:18 ` Andrew Morton
2012-11-21 8:30 ` Glauber Costa
2012-11-21 8:30 ` Glauber Costa
2012-11-12 12:19 ` kswapd0: excessive CPU usage Mel Gorman
2012-11-12 12:19 ` Mel Gorman
2012-11-12 13:13 ` Zdenek Kabelac
2012-11-12 13:13 ` Zdenek Kabelac
2012-11-12 13:31 ` Mel Gorman
2012-11-12 13:31 ` Mel Gorman
2012-11-12 14:50 ` Zdenek Kabelac
2012-11-12 14:50 ` Zdenek Kabelac
2012-11-18 19:00 ` Zdenek Kabelac
2012-11-18 19:00 ` Zdenek Kabelac
2012-11-18 19:07 ` Jiri Slaby
2012-11-18 19:07 ` Jiri Slaby
2012-11-09 8:40 ` Mel Gorman
2012-11-09 8:40 ` Mel Gorman
2012-10-11 22:14 ` kswapd0: wxcessive " Andrew Morton
2012-10-11 22:14 ` Andrew Morton
2012-10-11 22:26 ` Jiri Slaby
2012-10-11 22:26 ` Jiri Slaby
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=5096999F.1040405@redhat.com \
--to=riel@redhat.com \
--cc=Valdis.Kletnieks@vt.edu \
--cc=akpm@linux-foundation.org \
--cc=fedora@leemhuis.info \
--cc=jirislaby@gmail.com \
--cc=jslaby@suse.cz \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mgorman@suse.de \
/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.