lkml.org 
[lkml]   [2006]   [Jun]   [9]   [last100]   RSS Feed
Views: [more markup]  [less markup]  [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
DateFri, 09 Jun 2006 16:08:03 +0800
FromWu Fengguang <>
Subject[PATCH 2/5] readahead: aggressive initial sizes
Set ra_expect_bytes and ra_thrash_bytes of default_backing_dev_info to
large numbers.

Large initial values are better, because
	- most systems don't have the danger of thrashing
	- most small files are read in whole
	- they both increase slowly and drop rapidly

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

--- linux-2.6.17-rc5-mm3.orig/mm/readahead.c
+++ linux-2.6.17-rc5-mm3/mm/readahead.c
@@ -32,6 +32,9 @@
  * Adaptive read-ahead parameters.
  */
 
+/* Default max read-ahead size for the initial method.  */
+#define INITIAL_RA_PAGES  DIV_ROUND_UP(128*1024, PAGE_CACHE_SIZE)
+
 /* In laptop mode, poll delayed look-ahead on every ## pages read. */
 #define LAPTOP_POLL_INTERVAL 16
 
@@ -127,9 +130,9 @@ EXPORT_SYMBOL(default_unplug_io_fn);
 
 struct backing_dev_info default_backing_dev_info = {
 	.ra_pages	= MAX_RA_PAGES,
-	.ra_pages0	= (128*1024) / PAGE_CACHE_SIZE,
-	.ra_expect_bytes = VM_MIN_READAHEAD*1024,
-	.ra_thrash_bytes = VM_MIN_READAHEAD*1024,
+	.ra_pages0	= INITIAL_RA_PAGES,
+	.ra_expect_bytes = INITIAL_RA_PAGES * PAGE_CACHE_SIZE,
+	.ra_thrash_bytes = MAX_RA_PAGES * PAGE_CACHE_SIZE,
 	.state		= 0,
 	.capabilities	= BDI_CAP_MAP_COPY,
 	.unplug_io_fn	= default_unplug_io_fn,
--
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2006-06-09 10:15    [from the cache]
©2003-2008