All of lore.kernel.org
 help / color / mirror / Atom feed
From: Wu Fengguang <wfg@mail.ustc.edu.cn>
To: Andrew Morton <akpm@osdl.org>
Cc: linux-kernel@vger.kernel.org, Wu Fengguang <wfg@mail.ustc.edu.cn>
Subject: [PATCH 18/32] readahead: initial method - thrashing guard size
Date: Sat, 27 May 2006 23:49:07 +0800	[thread overview]
Message-ID: <348745093.16246@ustc.edu.cn> (raw)
Message-ID: <20060527155134.715578802@localhost.localdomain> (raw)
In-Reply-To: 20060527154849.927021763@localhost.localdomain

[-- Attachment #1: readahead-method-initial-size-thrash.patch --]
[-- Type: text/plain, Size: 1607 bytes --]

backing_dev_info.ra_thrash_bytes is dynamicly updated to be a little above
the thrashing safe read-ahead size. It is used in the initial method where
the thrashing threshold for the particular reader is still unknown.

Signed-off-by: Wu Fengguang <wfg@mail.ustc.edu.cn>
---

 mm/readahead.c |   20 ++++++++++++++++++++
 1 files changed, 20 insertions(+)

--- linux-2.6.17-rc4-mm3.orig/mm/readahead.c
+++ linux-2.6.17-rc4-mm3/mm/readahead.c
@@ -796,6 +796,22 @@ out:
 }
 
 /*
+ * Update `backing_dev_info.ra_thrash_bytes' to be a _biased_ average of
+ * read-ahead sizes. Which makes it an a-bit-risky(*) estimation of the
+ * _minimal_ read-ahead thrashing threshold on the device.
+ *
+ * (*) Note that being a bit risky can _help_ overall performance.
+ */
+static inline void update_ra_thrash_bytes(struct backing_dev_info *bdi,
+						unsigned long ra_size)
+{
+	ra_size <<= PAGE_CACHE_SHIFT;
+	bdi->ra_thrash_bytes = (bdi->ra_thrash_bytes < ra_size) ?
+				(ra_size + bdi->ra_thrash_bytes * 1023) / 1024:
+				(ra_size + bdi->ra_thrash_bytes *    7) /    8;
+}
+
+/*
  * The node's accumulated aging activities.
  */
 static unsigned long node_readahead_aging(void)
@@ -1144,6 +1160,10 @@ state_based_readahead(struct address_spa
 	if (!adjust_rala(growth_limit, &ra_size, &la_size))
 		return 0;
 
+	/* ra_size in its _steady_ state reflects thrashing threshold */
+	if (page && ra_old + ra_old / 8 >= ra_size)
+		update_ra_thrash_bytes(mapping->backing_dev_info, ra_size);
+
 	ra_set_class(ra, RA_CLASS_STATE);
 	ra_set_index(ra, index, ra->readahead_index);
 	ra_set_size(ra, ra_size, la_size);

--

  parent reply	other threads:[~2006-05-27 15:54 UTC|newest]

Thread overview: 30+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2006-05-27 15:48 [PATCH 00/32] Adaptive readahead V14 Wu Fengguang
2006-05-27 15:48 ` Wu Fengguang
2006-05-27 17:29   ` Michael Tokarev
2006-05-28 12:08     ` Wu Fengguang
2006-05-28 12:08       ` Wu Fengguang
2006-05-28 19:23         ` Michael Tokarev
2006-05-29  3:01           ` Wu Fengguang
2006-05-29  3:01             ` Wu Fengguang
2006-05-30  9:23             ` Jens Axboe
2006-05-30 11:32               ` Wu Fengguang
2006-05-30 11:32                 ` Wu Fengguang
2006-05-30 12:29                 ` Jens Axboe
2006-05-30 14:34                   ` Wu Fengguang
2006-05-30 14:34                     ` Wu Fengguang
2006-05-27 15:48 ` [PATCH 01/32] readahead: kconfig options Wu Fengguang
2006-05-27 15:48   ` Wu Fengguang
2006-05-27 15:48 ` [PATCH 04/32] mm: introduce PG_readahead Wu Fengguang
2006-05-27 15:48   ` Wu Fengguang
2006-05-27 15:48 ` [PATCH 06/32] readahead: delay page release in do_generic_mapping_read() Wu Fengguang
2006-05-27 15:48   ` Wu Fengguang
2006-05-27 15:48 ` [PATCH 07/32] readahead: insert cond_resched() calls Wu Fengguang
2006-05-27 15:48   ` Wu Fengguang
2006-05-27 15:48 ` [PATCH 08/32] readahead: {MIN,MAX}_RA_PAGES Wu Fengguang
2006-05-27 15:48   ` Wu Fengguang
2006-05-27 15:48 ` [PATCH 09/32] readahead: events accounting Wu Fengguang
2006-05-27 15:48   ` Wu Fengguang
2006-05-27 15:48 ` [PATCH 10/32] readahead: rescue_pages() Wu Fengguang
2006-05-27 15:48   ` Wu Fengguang
2006-05-27 15:49 ` [PATCH 11/32] readahead: sysctl parameters Wu Fengguang
2006-05-27 15:49   ` Wu Fengguang
2006-05-27 15:49 ` [PATCH 14/32] readahead: state based method - routines Wu Fengguang
2006-05-27 15:49   ` Wu Fengguang
2006-05-27 15:49 ` [PATCH 15/32] readahead: state based method Wu Fengguang
2006-05-27 15:49   ` Wu Fengguang
2006-05-27 15:49 ` Wu Fengguang [this message]
2006-05-27 15:49   ` [PATCH 18/32] readahead: initial method - thrashing guard size Wu Fengguang
2006-05-27 15:49 ` [PATCH 20/32] readahead: initial method - user recommended size Wu Fengguang
2006-05-27 15:49   ` Wu Fengguang
2006-05-27 15:49 ` [PATCH 22/32] readahead: backward prefetching method Wu Fengguang
2006-05-27 15:49   ` Wu Fengguang
2006-05-27 15:49 ` [PATCH 24/32] readahead: thrashing recovery method Wu Fengguang
2006-05-27 15:49   ` Wu Fengguang
2006-05-27 22:04     ` [PATCH 23/32] readahead: seeking reads method Ingo Oeser
2006-05-27 15:49 ` [PATCH 25/32] readahead: call scheme Wu Fengguang
2006-05-27 15:49   ` Wu Fengguang
2006-05-27 15:49 ` [PATCH 26/32] readahead: laptop mode Wu Fengguang
2006-05-27 15:49   ` Wu Fengguang
2006-05-27 15:49 ` [PATCH 27/32] readahead: loop case Wu Fengguang
2006-05-27 15:49   ` Wu Fengguang
2006-05-27 15:49 ` [PATCH 30/32] readahead: debug radix tree new functions Wu Fengguang
2006-05-27 15:49   ` Wu Fengguang
2006-05-27 15:49 ` [PATCH 31/32] readahead: debug traces showing accessed file names Wu Fengguang
2006-05-27 15:49   ` Wu Fengguang
2006-05-27 15:49 ` [PATCH 32/32] readahead: debug traces showing read patterns Wu Fengguang
2006-05-27 15:49   ` 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=348745093.16246@ustc.edu.cn \
    --to=wfg@mail.ustc.edu.cn \
    --cc=akpm@osdl.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 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.