Aggressive readahead policy for read on start-of-file.

Instead of selecting a conservative readahead size,
it tries to do large readahead in the first place.

However we have to watch on two cases:
	- do not ruin the hit rate for file-head-checkers
	- do not lead to thrashing for memory tight systems

In most cases, the threat of file-head-peakers is neglectable. So a simple
user tunable parameter is used. Users can try a larger initial read-ahead
size:
	echo 128 > /sys/block/sda/queue/initial_ra_kb

It benefits the many-small-files case:

adaptive readahead: avg 10.3 seconds
====================================

stock readahead: avg 12.3 seconds
=================================

The rxvt-unicode-7.7 being benchmarked is a dir with many .C/.h/.o files.

Signed-off-by: Wu Fengguang <wfg@mail.ustc.edu.cn>
Signed-off-by: Andrew Morton <akpm@osdl.org>
--- linux-2.6.19-rc5-mm2.orig/mm/readahead.c
+++ linux-2.6.19-rc5-mm2/mm/readahead.c
@@ -1440,6 +1440,42 @@ has_history_pages:
 }
 
 /*
+ * Read-ahead on start of file.
+ *
+ * We want to be as aggressive as possible, _and_
+ * 	- do not ruin the hit rate for file-head-peekers
+ * 	- do not lead to thrashing for memory tight systems
+ */
+static unsigned long
+initial_readahead(struct address_space *mapping, struct file *filp,
+		struct file_ra_state *ra, unsigned long req_size)
+{
+	struct backing_dev_info *bdi = mapping->backing_dev_info;
+	unsigned long thrash_pages = bdi->ra_thrash_bytes >> PAGE_CACHE_SHIFT;
+	unsigned long ra_size;
+	unsigned long la_size;
+
+	ra_size = req_size;
+
+	/* user imposed minimal size */
+	if (ra_size < bdi->ra_pages0)
+		ra_size = bdi->ra_pages0;
+
+	/* no read-ahead thrashing */
+	if (ra_size > thrash_pages)
+		ra_size = thrash_pages;
+
+	/* do look-ahead on large(>= 32KB) read-ahead */
+	la_size = ra_size / LOOKAHEAD_RATIO;
+
+	ra_set_class(ra, RA_CLASS_INITIAL);
+	ra_set_index(ra, 0, 0);
+	ra_set_size(ra, ra_size, la_size);
+
+	return ra_submit(ra, mapping, filp);
+}
+
+/*
  * ra_min is mainly determined by the size of cache memory. Reasonable?
  *
  * Table of concrete numbers for 4KB page size:

--
-
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/