linux-fsdevel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Jens Axboe <jens.axboe@oracle.com>
To: linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org,
	chris.mason@oracle.com, david@fromorbit.com, hch@infradead.org,
	akpm@linux-foundation.org, jack@suse.cz
Cc: Jens Axboe <jens.axboe@oracle.com>
Subject: [PATCH 06/10] writeback: allow sleepy exit of default writeback task
Date: Mon, 31 Aug 2009 14:14:47 +0200	[thread overview]
Message-ID: <1251720891-19793-7-git-send-email-jens.axboe@oracle.com> (raw)
In-Reply-To: <1251720891-19793-1-git-send-email-jens.axboe@oracle.com>

Since we do lazy create of default writeback tasks for a bdi, we can
allow sleepy exit if it has been completely idle for 5 minutes.

Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
---
 fs/fs-writeback.c           |   30 +++++++++++++++++++++++++-----
 include/linux/backing-dev.h |    5 +++++
 include/linux/writeback.h   |    2 +-
 3 files changed, 31 insertions(+), 6 deletions(-)

diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c
index 0423880..7e139d5 100644
--- a/fs/fs-writeback.c
+++ b/fs/fs-writeback.c
@@ -373,7 +373,7 @@ static struct bdi_work *get_next_work_item(struct backing_dev_info *bdi,
 /*
  * Retrieve work items and do the writeback they describe
  */
-void wb_do_writeback(struct bdi_writeback *wb, int force_wait)
+long wb_do_writeback(struct bdi_writeback *wb, int force_wait)
 {
 	struct backing_dev_info *bdi = wb->bdi;
 	struct bdi_work *work;
@@ -417,8 +417,10 @@ void wb_do_writeback(struct bdi_writeback *wb, int force_wait)
 				global_page_state(NR_UNSTABLE_NFS) +
 				(inodes_stat.nr_inodes - inodes_stat.nr_unused);
 
-		wb_writeback(wb, nr_pages, NULL, WB_SYNC_NONE, 1);
+		wrote = wb_writeback(wb, nr_pages, NULL, WB_SYNC_NONE, 1);
 	}
+
+	return wrote;
 }
 
 /*
@@ -427,10 +429,28 @@ void wb_do_writeback(struct bdi_writeback *wb, int force_wait)
  */
 int bdi_writeback_task(struct bdi_writeback *wb)
 {
+	unsigned long last_active = jiffies;
+	unsigned long wait_jiffies = -1UL;
+	long pages_written;
+
 	while (!kthread_should_stop()) {
-		unsigned long wait_jiffies;
+		pages_written = wb_do_writeback(wb, 0);
+
+		if (pages_written)
+			last_active = jiffies;
+		else if (wait_jiffies != -1UL) {
+			unsigned long max_idle;
 
-		wb_do_writeback(wb, 0);
+			/*
+			 * Longest period of inactivity that we tolerate. If we
+			 * see dirty data again later, the task will get
+			 * recreated automatically.
+			 */
+			max_idle = max(5UL * 60 * HZ, wait_jiffies);
+			if (time_after(jiffies, max_idle + last_active) &&
+			    wb_is_default_task(wb))
+				break;
+		}
 
 		wait_jiffies = msecs_to_jiffies(dirty_writeback_interval * 10);
 		set_current_state(TASK_INTERRUPTIBLE);
diff --git a/include/linux/backing-dev.h b/include/linux/backing-dev.h
index 90a1910..7a23f2f 100644
--- a/include/linux/backing-dev.h
+++ b/include/linux/backing-dev.h
@@ -110,6 +110,11 @@ int bdi_has_dirty_io(struct backing_dev_info *bdi);
 extern spinlock_t bdi_lock;
 extern struct list_head bdi_list;
 
+static inline int wb_is_default_task(struct bdi_writeback *wb)
+{
+	return wb == &wb->bdi->wb;
+}
+
 static inline int bdi_wblist_needs_lock(struct backing_dev_info *bdi)
 {
 	return test_bit(BDI_wblist_lock, &bdi->state);
diff --git a/include/linux/writeback.h b/include/linux/writeback.h
index 588a449..e070b91 100644
--- a/include/linux/writeback.h
+++ b/include/linux/writeback.h
@@ -68,7 +68,7 @@ struct writeback_control {
 void writeback_inodes(struct writeback_control *wbc);
 int inode_wait(void *);
 void sync_inodes_sb(struct super_block *, int wait);
-void wb_do_writeback(struct bdi_writeback *wb, int force_wait);
+long wb_do_writeback(struct bdi_writeback *wb, int force_wait);
 
 /* writeback.h requires fs.h; it, too, is not included from here. */
 static inline void wait_on_inode(struct inode *inode)
-- 
1.6.4.1.207.g68ea

  parent reply	other threads:[~2009-08-31 12:14 UTC|newest]

Thread overview: 21+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-08-31 12:14 [PATCH 0/10] Per-bdi writeback flusher threads v14 Jens Axboe
2009-08-31 12:14 ` [PATCH 01/10] writeback: move dirty inodes from super_block to backing_dev_info Jens Axboe
2009-08-31 12:14 ` [PATCH 02/10] writeback: switch to per-bdi threads for flushing data Jens Axboe
2009-08-31 12:58   ` Christoph Hellwig
2009-08-31 17:29     ` Jens Axboe
2009-08-31 12:14 ` [PATCH 03/10] writeback: get rid of pdflush completely Jens Axboe
2009-08-31 12:59   ` Christoph Hellwig
2009-08-31 17:29     ` Jens Axboe
2009-08-31 12:14 ` [PATCH 04/10] writeback: separate the flushing state/task from the bdi Jens Axboe
2009-08-31 12:14 ` [PATCH 05/10] writeback: support > 1 flusher thread per bdi Jens Axboe
2009-08-31 13:01   ` Christoph Hellwig
2009-08-31 17:32     ` Jens Axboe
2009-08-31 12:14 ` Jens Axboe [this message]
2009-08-31 13:02   ` [PATCH 06/10] writeback: allow sleepy exit of default writeback task Christoph Hellwig
2009-08-31 17:32     ` Jens Axboe
2009-08-31 12:14 ` [PATCH 07/10] writeback: add some debug inode list counters to bdi stats Jens Axboe
2009-08-31 12:14 ` [PATCH 08/10] writeback: add name to backing_dev_info Jens Axboe
2009-08-31 12:14 ` [PATCH 09/10] writeback: check for registered bdi in flusher add and inode dirty Jens Axboe
2009-08-31 12:14 ` [PATCH 10/10] vm: Add an tuning knob for vm.max_writeback_pages Jens Axboe
  -- strict thread matches above, loose matches on Subject: below --
2009-06-25 10:41 [PATCH 0/10] Per-bdi writeback flusher threads v12 Jens Axboe
2009-06-25 10:41 ` [PATCH 06/10] writeback: allow sleepy exit of default writeback task Jens Axboe
2009-06-17 11:07 [PATCH 0/10] Per-bdi writeback flusher threads v11 Jens Axboe
2009-06-17 11:07 ` [PATCH 06/10] writeback: allow sleepy exit of default writeback task Jens Axboe

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=1251720891-19793-7-git-send-email-jens.axboe@oracle.com \
    --to=jens.axboe@oracle.com \
    --cc=akpm@linux-foundation.org \
    --cc=chris.mason@oracle.com \
    --cc=david@fromorbit.com \
    --cc=hch@infradead.org \
    --cc=jack@suse.cz \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    /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).