All of lore.kernel.org
 help / color / mirror / Atom feed
From: Marcin Slusarz <marcin.slusarz@gmail.com>
To: Christoph Lameter <cl@linux-foundation.org>,
	Pekka Enberg <penberg@kernel.org>, Matt Mackall <mpm@selenic.com>
Cc: LKML <linux-kernel@vger.kernel.org>, linux-mm@kvack.org
Subject: [PATCH] slub: reduce overhead of slub_debug
Date: Sun, 26 Jun 2011 21:39:18 +0200	[thread overview]
Message-ID: <20110626193918.GA3339@joi.lan> (raw)

slub checks for poison one byte by one, which is highly inefficient
and shows up frequently as a highest cpu-eater in perf top.

Joining reads gives nice speedup:

(Compiling some project with different options)
                                 make -j12    make clean
slub_debug disabled:             1m 27s       1.2 s
slub_debug enabled:              1m 46s       7.6 s
slub_debug enabled + this patch: 1m 33s       3.2 s

check_bytes still shows up high, but not always at the top.

Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Cc: Christoph Lameter <cl@linux-foundation.org>
Cc: Pekka Enberg <penberg@kernel.org>
Cc: Matt Mackall <mpm@selenic.com>
Cc: linux-mm@kvack.org
---
 mm/slub.c |   36 ++++++++++++++++++++++++++++++++++--
 1 files changed, 34 insertions(+), 2 deletions(-)

diff --git a/mm/slub.c b/mm/slub.c
index 35f351f..a40ef2d 100644
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -557,10 +557,10 @@ static void init_object(struct kmem_cache *s, void *object, u8 val)
 		memset(p + s->objsize, val, s->inuse - s->objsize);
 }
 
-static u8 *check_bytes(u8 *start, unsigned int value, unsigned int bytes)
+static u8 *check_bytes8(u8 *start, u8 value, unsigned int bytes)
 {
 	while (bytes) {
-		if (*start != (u8)value)
+		if (*start != value)
 			return start;
 		start++;
 		bytes--;
@@ -568,6 +568,38 @@ static u8 *check_bytes(u8 *start, unsigned int value, unsigned int bytes)
 	return NULL;
 }
 
+static u8 *check_bytes(u8 *start, u8 value, unsigned int bytes)
+{
+	u64 value64;
+	unsigned int words, prefix;
+
+	if (bytes <= 16)
+		return check_bytes8(start, value, bytes);
+
+	value64 = value | value << 8 | value << 16 | value << 24;
+	value64 = value64 | value64 << 32;
+	prefix = 8 - ((unsigned long)start) % 8;
+
+	if (prefix) {
+		u8 *r = check_bytes8(start, value, prefix);
+		if (r)
+			return r;
+		start += prefix;
+		bytes -= prefix;
+	}
+
+	words = bytes / 8;
+
+	while (words) {
+		if (*(u64 *)start != value64)
+			return check_bytes8(start, value, 8);
+		start += 8;
+		words--;
+	}
+
+	return check_bytes8(start, value, bytes % 8);
+}
+
 static void restore_bytes(struct kmem_cache *s, char *message, u8 data,
 						void *from, void *to)
 {
-- 
1.7.5.3


WARNING: multiple messages have this Message-ID (diff)
From: Marcin Slusarz <marcin.slusarz@gmail.com>
To: Christoph Lameter <cl@linux-foundation.org>,
	Pekka Enberg <penberg@kernel.org>, Matt Mackall <mpm@selenic.com>
Cc: LKML <linux-kernel@vger.kernel.org>, linux-mm@kvack.org
Subject: [PATCH] slub: reduce overhead of slub_debug
Date: Sun, 26 Jun 2011 21:39:18 +0200	[thread overview]
Message-ID: <20110626193918.GA3339@joi.lan> (raw)

slub checks for poison one byte by one, which is highly inefficient
and shows up frequently as a highest cpu-eater in perf top.

Joining reads gives nice speedup:

(Compiling some project with different options)
                                 make -j12    make clean
slub_debug disabled:             1m 27s       1.2 s
slub_debug enabled:              1m 46s       7.6 s
slub_debug enabled + this patch: 1m 33s       3.2 s

check_bytes still shows up high, but not always at the top.

Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Cc: Christoph Lameter <cl@linux-foundation.org>
Cc: Pekka Enberg <penberg@kernel.org>
Cc: Matt Mackall <mpm@selenic.com>
Cc: linux-mm@kvack.org
---
 mm/slub.c |   36 ++++++++++++++++++++++++++++++++++--
 1 files changed, 34 insertions(+), 2 deletions(-)

diff --git a/mm/slub.c b/mm/slub.c
index 35f351f..a40ef2d 100644
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -557,10 +557,10 @@ static void init_object(struct kmem_cache *s, void *object, u8 val)
 		memset(p + s->objsize, val, s->inuse - s->objsize);
 }
 
-static u8 *check_bytes(u8 *start, unsigned int value, unsigned int bytes)
+static u8 *check_bytes8(u8 *start, u8 value, unsigned int bytes)
 {
 	while (bytes) {
-		if (*start != (u8)value)
+		if (*start != value)
 			return start;
 		start++;
 		bytes--;
@@ -568,6 +568,38 @@ static u8 *check_bytes(u8 *start, unsigned int value, unsigned int bytes)
 	return NULL;
 }
 
+static u8 *check_bytes(u8 *start, u8 value, unsigned int bytes)
+{
+	u64 value64;
+	unsigned int words, prefix;
+
+	if (bytes <= 16)
+		return check_bytes8(start, value, bytes);
+
+	value64 = value | value << 8 | value << 16 | value << 24;
+	value64 = value64 | value64 << 32;
+	prefix = 8 - ((unsigned long)start) % 8;
+
+	if (prefix) {
+		u8 *r = check_bytes8(start, value, prefix);
+		if (r)
+			return r;
+		start += prefix;
+		bytes -= prefix;
+	}
+
+	words = bytes / 8;
+
+	while (words) {
+		if (*(u64 *)start != value64)
+			return check_bytes8(start, value, 8);
+		start += 8;
+		words--;
+	}
+
+	return check_bytes8(start, value, bytes % 8);
+}
+
 static void restore_bytes(struct kmem_cache *s, char *message, u8 data,
 						void *from, void *to)
 {
-- 
1.7.5.3

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

             reply	other threads:[~2011-06-26 19:41 UTC|newest]

Thread overview: 44+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-06-26 19:39 Marcin Slusarz [this message]
2011-06-26 19:39 ` [PATCH] slub: reduce overhead of slub_debug Marcin Slusarz
2011-06-28 19:32 ` Christoph Lameter
2011-06-28 19:32   ` Christoph Lameter
2011-06-28 19:40   ` David Daney
2011-06-28 19:40     ` David Daney
2011-06-28 20:58     ` David Rientjes
2011-06-28 20:58       ` David Rientjes
2011-06-28 21:04       ` Ben Greear
2011-06-28 21:04         ` Ben Greear
2011-06-28 21:10         ` David Rientjes
2011-06-28 21:10           ` David Rientjes
2011-06-28 21:16       ` Dave Jones
2011-06-28 21:16         ` Dave Jones
2011-07-07 18:07 ` Pekka Enberg
2011-07-07 18:07   ` Pekka Enberg
2011-07-07 18:17   ` Christoph Lameter
2011-07-07 18:17     ` Christoph Lameter
2011-07-07 18:30     ` Ben Greear
2011-07-07 18:30       ` Ben Greear
2011-07-07 18:42       ` Christoph Lameter
2011-07-07 18:42         ` Christoph Lameter
2011-07-07 18:54         ` Ben Greear
2011-07-07 18:54           ` Ben Greear
2011-07-07 18:30     ` Matt Mackall
2011-07-07 18:30       ` Matt Mackall
2011-07-07 18:52     ` Pekka Enberg
2011-07-07 18:52       ` Pekka Enberg
2011-07-07 18:55       ` Matt Mackall
2011-07-07 18:55         ` Matt Mackall
2011-07-07 19:12       ` Christoph Lameter
2011-07-07 19:12         ` Christoph Lameter
2011-07-07 19:21         ` David Miller
2011-07-07 19:21           ` David Miller
2011-07-07 19:49           ` Pekka Enberg
2011-07-07 19:49             ` Pekka Enberg
2011-07-07 20:12             ` Christoph Lameter
2011-07-07 20:12               ` Christoph Lameter
2011-07-08  5:23               ` Andi Kleen
2011-07-08  5:23                 ` Andi Kleen
2011-07-08 17:41                 ` Christoph Lameter
2011-07-08 17:41                   ` Christoph Lameter
2011-07-08  5:38               ` Pekka Enberg
2011-07-08  5:38                 ` Pekka Enberg

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=20110626193918.GA3339@joi.lan \
    --to=marcin.slusarz@gmail.com \
    --cc=cl@linux-foundation.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=mpm@selenic.com \
    --cc=penberg@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.