From: Mel Gorman <mel@csn.ul.ie>
To: linux-mm@kvack.org, linux-fsdevel@vger.kernel.org
Cc: Mel Gorman <mel@csn.ul.ie>,
Andrew Morton <akpm@linux-foundation.org>,
Christian Ehrhardt <ehrhardt@linux.vnet.ibm.com>,
Johannes Weiner <hannes@cmpxchg.org>,
Wu Fengguang <fengguang.wu@intel.com>, Jan Kara <jack@suse.cz>,
linux-kernel@vger.kernel.org
Subject: [PATCH 2/3] writeback: Record if the congestion was unnecessary
Date: Thu, 26 Aug 2010 16:14:15 +0100 [thread overview]
Message-ID: <1282835656-5638-3-git-send-email-mel@csn.ul.ie> (raw)
In-Reply-To: <1282835656-5638-1-git-send-email-mel@csn.ul.ie>
If congestion_wait() is called when there is no congestion, the caller
will wait for the full timeout. This can cause unreasonable and
unnecessary stalls. There are a number of potential modifications that
could be made to wake sleepers but this patch measures how serious the
problem is. It keeps count of how many congested BDIs there are. If
congestion_wait() is called with no BDIs congested, the tracepoint will
record that the wait was unnecessary.
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
---
include/trace/events/writeback.h | 11 ++++++++---
mm/backing-dev.c | 15 ++++++++++++---
2 files changed, 20 insertions(+), 6 deletions(-)
diff --git a/include/trace/events/writeback.h b/include/trace/events/writeback.h
index e3bee61..03bb04b 100644
--- a/include/trace/events/writeback.h
+++ b/include/trace/events/writeback.h
@@ -155,19 +155,24 @@ DEFINE_WBC_EVENT(wbc_writepage);
TRACE_EVENT(writeback_congest_waited,
- TP_PROTO(unsigned int usec_delayed),
+ TP_PROTO(unsigned int usec_delayed, bool unnecessary),
- TP_ARGS(usec_delayed),
+ TP_ARGS(usec_delayed, unnecessary),
TP_STRUCT__entry(
__field( unsigned int, usec_delayed )
+ __field( unsigned int, unnecessary )
),
TP_fast_assign(
__entry->usec_delayed = usec_delayed;
+ __entry->unnecessary = unnecessary;
),
- TP_printk("usec_delayed=%u", __entry->usec_delayed)
+ TP_printk("usec_delayed=%u unnecessary=%d",
+ __entry->usec_delayed,
+ __entry->unnecessary
+ )
);
#endif /* _TRACE_WRITEBACK_H */
diff --git a/mm/backing-dev.c b/mm/backing-dev.c
index 7ae33e2..a49167f 100644
--- a/mm/backing-dev.c
+++ b/mm/backing-dev.c
@@ -724,6 +724,7 @@ static wait_queue_head_t congestion_wqh[2] = {
__WAIT_QUEUE_HEAD_INITIALIZER(congestion_wqh[0]),
__WAIT_QUEUE_HEAD_INITIALIZER(congestion_wqh[1])
};
+static atomic_t nr_bdi_congested[2];
void clear_bdi_congested(struct backing_dev_info *bdi, int sync)
{
@@ -731,7 +732,8 @@ void clear_bdi_congested(struct backing_dev_info *bdi, int sync)
wait_queue_head_t *wqh = &congestion_wqh[sync];
bit = sync ? BDI_sync_congested : BDI_async_congested;
- clear_bit(bit, &bdi->state);
+ if (test_and_clear_bit(bit, &bdi->state))
+ atomic_dec(&nr_bdi_congested[sync]);
smp_mb__after_clear_bit();
if (waitqueue_active(wqh))
wake_up(wqh);
@@ -743,7 +745,8 @@ void set_bdi_congested(struct backing_dev_info *bdi, int sync)
enum bdi_state bit;
bit = sync ? BDI_sync_congested : BDI_async_congested;
- set_bit(bit, &bdi->state);
+ if (!test_and_set_bit(bit, &bdi->state))
+ atomic_inc(&nr_bdi_congested[sync]);
}
EXPORT_SYMBOL(set_bdi_congested);
@@ -760,14 +763,20 @@ long congestion_wait(int sync, long timeout)
{
long ret;
unsigned long start = jiffies;
+ bool unnecessary = false;
DEFINE_WAIT(wait);
wait_queue_head_t *wqh = &congestion_wqh[sync];
+ /* Check if this call to congestion_wait was necessary */
+ if (atomic_read(&nr_bdi_congested[sync]) == 0)
+ unnecessary = true;
+
prepare_to_wait(wqh, &wait, TASK_UNINTERRUPTIBLE);
ret = io_schedule_timeout(timeout);
finish_wait(wqh, &wait);
- trace_writeback_congest_waited(jiffies_to_usecs(jiffies - start));
+ trace_writeback_congest_waited(jiffies_to_usecs(jiffies - start),
+ unnecessary);
return ret;
}
--
1.7.1
next prev parent reply other threads:[~2010-08-26 15:14 UTC|newest]
Thread overview: 38+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-08-26 15:14 [RFC PATCH 0/3] Do not wait the full timeout on congestion_wait when there is no congestion Mel Gorman
2010-08-26 15:14 ` [PATCH 1/3] writeback: Account for time spent congestion_waited Mel Gorman
2010-08-26 17:23 ` Minchan Kim
2010-08-26 18:10 ` Johannes Weiner
2010-08-26 15:14 ` Mel Gorman [this message]
2010-08-26 17:35 ` [PATCH 2/3] writeback: Record if the congestion was unnecessary Minchan Kim
2010-08-26 17:41 ` Mel Gorman
2010-08-26 18:29 ` Johannes Weiner
2010-08-26 20:31 ` Mel Gorman
2010-08-27 2:12 ` Shaohua Li
2010-08-27 9:20 ` Mel Gorman
2010-08-27 8:16 ` Johannes Weiner
2010-08-27 9:24 ` Mel Gorman
2010-08-30 13:19 ` Johannes Weiner
2010-08-31 15:02 ` Mel Gorman
2010-09-02 15:49 ` Johannes Weiner
2010-09-02 18:28 ` Mel Gorman
2010-08-29 16:03 ` Minchan Kim
2010-08-26 15:14 ` [PATCH 3/3] writeback: Do not congestion sleep when there are no congested BDIs Mel Gorman
2010-08-26 17:38 ` Minchan Kim
2010-08-26 17:42 ` Mel Gorman
2010-08-26 18:17 ` Johannes Weiner
2010-08-26 20:23 ` Mel Gorman
2010-08-27 1:11 ` Wu Fengguang
2010-08-27 9:34 ` Mel Gorman
2010-08-27 1:42 ` Wu Fengguang
2010-08-27 9:37 ` Mel Gorman
2010-08-27 5:13 ` Dave Chinner
2010-08-27 9:33 ` Mel Gorman
2010-08-26 17:20 ` [RFC PATCH 0/3] Do not wait the full timeout on congestion_wait when there is no congestion Minchan Kim
2010-08-26 17:31 ` Mel Gorman
2010-08-26 17:50 ` Minchan Kim
2010-08-27 1:21 ` Wu Fengguang
2010-08-27 1:41 ` Minchan Kim
2010-08-27 1:50 ` Wu Fengguang
2010-08-27 2:02 ` Minchan Kim
2010-08-27 4:34 ` Wu Fengguang
2010-08-27 9:38 ` Mel Gorman
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=1282835656-5638-3-git-send-email-mel@csn.ul.ie \
--to=mel@csn.ul.ie \
--cc=akpm@linux-foundation.org \
--cc=ehrhardt@linux.vnet.ibm.com \
--cc=fengguang.wu@intel.com \
--cc=hannes@cmpxchg.org \
--cc=jack@suse.cz \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.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).