linux-fsdevel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Wu Fengguang <fengguang.wu@intel.com>
To: Andrew Morton <akpm@linux-foundation.org>
Cc: Jan Kara <jack@suse.cz>, Dave Chinner <david@fromorbit.com>,
	Wu Fengguang <fengguang.wu@intel.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: Theodore Ts'o <tytso@mit.edu>
Cc: Chris Mason <chris.mason@oracle.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Mel Gorman <mel@csn.ul.ie>
Cc: Rik van Riel <riel@redhat.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Greg Thelen <gthelen@google.com>
Cc: Minchan Kim <minchan.kim@gmail.com>
Cc: linux-mm <linux-mm@kvack.org>
Cc: <linux-fsdevel@vger.kernel.org>
Cc: LKML <linux-kernel@vger.kernel.org>
Subject: [PATCH 16/35] writeback: increase min pause time on concurrent dirtiers
Date: Mon, 13 Dec 2010 22:47:02 +0800	[thread overview]
Message-ID: <20101213150328.284979629@intel.com> (raw)
In-Reply-To: 20101213144646.341970461@intel.com

[-- Attachment #1: writeback-min-pause-time-for-concurrent-dirtiers.patch --]
[-- Type: text/plain, Size: 2400 bytes --]

Target for >60ms pause time when there are 100+ heavy dirtiers per bdi.
(will average around 100ms given 200ms max pause time)

It's OK for 1 dd task doing 100MB/s to be throttle paused 100 times per
second.  However when there are 100 tasks writing to the same disk,
That sums up to 100*100 balance_dirty_pages() calls per second and may
lead to massive cacheline bouncing on accessing the global page states
in NUMA machines.  Even in single socket boxes, we easily see >10% CPU
time reduction by increasing the pause time.

CC: Dave Chinner <david@fromorbit.com>
Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
---
 mm/page-writeback.c |   23 ++++++++++++++++++++++-
 1 file changed, 22 insertions(+), 1 deletion(-)

--- linux-next.orig/mm/page-writeback.c	2010-12-13 21:46:16.000000000 +0800
+++ linux-next/mm/page-writeback.c	2010-12-13 21:46:16.000000000 +0800
@@ -659,6 +659,27 @@ static unsigned long max_pause(unsigned 
 }
 
 /*
+ * Scale up pause time for concurrent dirtiers in order to reduce CPU overheads.
+ * But ensure reasonably large [min_pause, max_pause] range size, so that
+ * nr_dirtied_pause (and hence future pause time) can stay reasonably stable.
+ */
+static unsigned long min_pause(struct backing_dev_info *bdi,
+			       unsigned long max)
+{
+	unsigned long hi = ilog2(bdi->write_bandwidth);
+	unsigned long lo = ilog2(bdi->throttle_bandwidth);
+	unsigned long t;
+
+	if (lo >= hi)
+		return 1;
+
+	/* (N * 10ms) on 2^N concurrent tasks */
+	t = (hi - lo) * (10 * HZ) / 1024;
+
+	return clamp_val(t, 1, max / 2);
+}
+
+/*
  * balance_dirty_pages() must be called by processes which are generating dirty
  * data.  It looks at the number of dirty pages in the machine and will force
  * the caller to perform writeback if the system is over `vm_dirty_ratio'.
@@ -798,7 +819,7 @@ pause:
 
 	if (pause == 0 && nr_dirty < background_thresh)
 		current->nr_dirtied_pause = ratelimit_pages(bdi);
-	else if (pause == 1)
+	else if (pause <= min_pause(bdi, pause_max))
 		current->nr_dirtied_pause += current->nr_dirtied_pause / 32 + 1;
 	else if (pause >= pause_max)
 		/*


--
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 policy in Canada: sign http://dissolvethecrtc.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

  parent reply	other threads:[~2010-12-13 14:47 UTC|newest]

Thread overview: 87+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-12-13 14:46 [PATCH 00/35] IO-less dirty throttling v4 Wu Fengguang
2010-12-13 14:46 ` [PATCH 01/35] writeback: enabling gate limit for light dirtied bdi Wu Fengguang
2011-01-12 21:43   ` Jan Kara
2011-01-13  3:44     ` Wu Fengguang
2011-01-13  3:58       ` Wu Fengguang
2011-01-13 19:26       ` Peter Zijlstra
2011-01-14  3:21         ` Wu Fengguang
2010-12-13 14:46 ` [PATCH 02/35] writeback: safety margin for bdi stat error Wu Fengguang
2011-01-12 21:59   ` Jan Kara
2011-01-13  4:14     ` Wu Fengguang
2011-01-13 10:38       ` Jan Kara
2011-01-13 10:41         ` Wu Fengguang
2010-12-13 14:46 ` [PATCH 03/35] writeback: prevent duplicate balance_dirty_pages_ratelimited() calls Wu Fengguang
2010-12-13 14:46 ` [PATCH 04/35] writeback: reduce per-bdi dirty threshold ramp up time Wu Fengguang
2010-12-14 13:37   ` Richard Kennedy
2010-12-14 13:59     ` Wu Fengguang
2010-12-14 14:33       ` Wu Fengguang
2010-12-14 14:39         ` Wu Fengguang
2010-12-14 14:50           ` Peter Zijlstra
2010-12-14 15:15             ` Wu Fengguang
2010-12-14 15:26               ` Wu Fengguang
2010-12-14 14:56           ` Wu Fengguang
2010-12-15 18:48       ` Richard Kennedy
2010-12-17 13:07         ` Wu Fengguang
2010-12-13 14:46 ` [PATCH 05/35] writeback: IO-less balance_dirty_pages() Wu Fengguang
2010-12-13 14:46 ` [PATCH 06/35] writeback: consolidate variable names in balance_dirty_pages() Wu Fengguang
2010-12-13 14:46 ` [PATCH 07/35] writeback: per-task rate limit on balance_dirty_pages() Wu Fengguang
2010-12-13 14:46 ` [PATCH 08/35] writeback: user space think time compensation Wu Fengguang
2010-12-13 14:46 ` [PATCH 09/35] writeback: account per-bdi accumulated written pages Wu Fengguang
2010-12-13 14:46 ` [PATCH 10/35] writeback: bdi write bandwidth estimation Wu Fengguang
2010-12-13 14:46 ` [PATCH 11/35] writeback: show bdi write bandwidth in debugfs Wu Fengguang
2010-12-13 14:46 ` [PATCH 12/35] writeback: scale down max throttle bandwidth on concurrent dirtiers Wu Fengguang
2010-12-14  1:21   ` Yan, Zheng
2010-12-14  7:00     ` Wu Fengguang
2010-12-14 13:00       ` Wu Fengguang
2010-12-13 14:46 ` [PATCH 13/35] writeback: bdi base throttle bandwidth Wu Fengguang
2010-12-13 14:47 ` [PATCH 14/35] writeback: smoothed bdi dirty pages Wu Fengguang
2010-12-13 14:47 ` [PATCH 15/35] writeback: adapt max balance pause time to memory size Wu Fengguang
2010-12-13 14:47 ` Wu Fengguang [this message]
2010-12-13 18:23   ` [PATCH 16/35] writeback: increase min pause time on concurrent dirtiers Valdis.Kletnieks
2010-12-14  6:51     ` Wu Fengguang
2010-12-14 18:42       ` Valdis.Kletnieks
2010-12-14 18:55         ` Peter Zijlstra
2010-12-14 20:13           ` Valdis.Kletnieks
2010-12-14 20:24             ` Peter Zijlstra
2010-12-14 20:37               ` Valdis.Kletnieks
2010-12-13 14:47 ` [PATCH 17/35] writeback: quit throttling when bdi dirty pages dropped low Wu Fengguang
2010-12-16  5:17   ` Wu Fengguang
2010-12-13 14:47 ` [PATCH 18/35] writeback: start background writeback earlier Wu Fengguang
2010-12-16  5:37   ` Wu Fengguang
2010-12-13 14:47 ` [PATCH 19/35] writeback: make nr_to_write a per-file limit Wu Fengguang
2010-12-13 14:47 ` [PATCH 20/35] writeback: scale IO chunk size up to device bandwidth Wu Fengguang
2010-12-13 14:47 ` [PATCH 21/35] writeback: trace balance_dirty_pages() Wu Fengguang
2010-12-13 14:47 ` [PATCH 22/35] writeback: trace global dirty page states Wu Fengguang
2010-12-17  2:19   ` Wu Fengguang
2010-12-17  3:11     ` Wu Fengguang
2010-12-17  6:52     ` Hugh Dickins
2010-12-17  9:31       ` Wu Fengguang
2010-12-17 11:21       ` [PATCH] writeback: skip balance_dirty_pages() for in-memory fs Wu Fengguang
2010-12-17 14:21         ` Rik van Riel
2010-12-17 15:34         ` Minchan Kim
2010-12-17 15:42           ` Minchan Kim
2010-12-21  5:59         ` Hugh Dickins
2010-12-21  9:39           ` Wu Fengguang
2010-12-30  3:15             ` Hugh Dickins
2010-12-13 14:47 ` [PATCH 23/35] writeback: trace writeback_single_inode() Wu Fengguang
2010-12-13 14:47 ` [PATCH 24/35] btrfs: dont call balance_dirty_pages_ratelimited() on already dirty pages Wu Fengguang
2010-12-13 14:47 ` [PATCH 25/35] btrfs: lower the dirty balacing rate limit Wu Fengguang
2010-12-13 14:47 ` [PATCH 26/35] btrfs: wait on too many nr_async_bios Wu Fengguang
2010-12-13 14:47 ` [PATCH 27/35] nfs: livelock prevention is now done in VFS Wu Fengguang
2010-12-13 14:47 ` [PATCH 28/35] nfs: writeback pages wait queue Wu Fengguang
2010-12-13 14:47 ` [PATCH 29/35] nfs: in-commit pages accounting and " Wu Fengguang
2010-12-13 21:15   ` Trond Myklebust
2010-12-14 15:40     ` Wu Fengguang
2010-12-14 15:57       ` Trond Myklebust
2010-12-15 15:07         ` Wu Fengguang
2010-12-13 14:47 ` [PATCH 30/35] nfs: heuristics to avoid commit Wu Fengguang
2010-12-13 20:53   ` Trond Myklebust
2010-12-14  8:20     ` Wu Fengguang
2010-12-13 14:47 ` [PATCH 31/35] nfs: dont change wbc->nr_to_write in write_inode() Wu Fengguang
2010-12-13 21:01   ` Trond Myklebust
2010-12-14 15:53     ` Wu Fengguang
2010-12-13 14:47 ` [PATCH 32/35] nfs: limit the range of commits Wu Fengguang
2010-12-13 14:47 ` [PATCH 33/35] nfs: adapt congestion threshold to dirty threshold Wu Fengguang
2010-12-13 14:47 ` [PATCH 34/35] nfs: trace nfs_commit_unstable_pages() Wu Fengguang
2010-12-13 14:47 ` [PATCH 35/35] nfs: trace nfs_commit_release() Wu Fengguang
     [not found] ` <AANLkTinFeu7LMaDFgUcP3r2oqVHE5bei3T5JTPGBNvS9@mail.gmail.com>
2010-12-14  4:59   ` [PATCH 00/35] IO-less dirty throttling v4 Wu Fengguang

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=20101213150328.284979629@intel.com \
    --to=fengguang.wu@intel.com \
    --cc=akpm@linux-foundation.org \
    --cc=david@fromorbit.com \
    --cc=jack@suse.cz \
    /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).