From: Christoph Lameter <cl@linux.com>
To: Pekka Enberg <penberg@cs.helsinki.fi>
Cc: David Rientjes <rientjes@google.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: linux-kernel@vger.kernel.org
Subject: [slubllv3 11/21] slub: Add cmpxchg_double_slab()
Date: Fri, 15 Apr 2011 15:47:41 -0500 [thread overview]
Message-ID: <20110415204752.445029138@linux.com> (raw)
In-Reply-To: 20110415204730.326790555@linux.com
[-- Attachment #1: cmpxchg_double_slab --]
[-- Type: text/plain, Size: 3794 bytes --]
Add a function that operates on the second doubleword in the page struct
and manipulates the object counters, the freelist and the frozen attribute.
Signed-off-by: Christoph Lameter <cl@linux.com>
---
include/linux/slub_def.h | 1 +
mm/slub.c | 45 +++++++++++++++++++++++++++++++++++++++++++++
2 files changed, 46 insertions(+)
Index: linux-2.6/mm/slub.c
===================================================================
--- linux-2.6.orig/mm/slub.c 2011-04-15 13:14:40.000000000 -0500
+++ linux-2.6/mm/slub.c 2011-04-15 13:14:51.000000000 -0500
@@ -131,6 +131,9 @@ static inline int kmem_cache_debug(struc
/* Enable to test recovery from slab corruption on boot */
#undef SLUB_RESILIENCY_TEST
+/* Enable to log cmpxchg failures */
+#undef SLUB_DEBUG_CMPXCHG
+
/*
* Mininum number of partial slabs. These will be left on the partial
* lists even if they are empty. kmem_cache_shrink may reclaim them.
@@ -170,6 +173,7 @@ static inline int kmem_cache_debug(struc
/* Internal SLUB flags */
#define __OBJECT_POISON 0x80000000UL /* Poison object */
+#define __CMPXCHG_DOUBLE 0x40000000UL /* Use cmpxchg_double */
static int kmem_size = sizeof(struct kmem_cache);
@@ -326,6 +330,37 @@ static inline int oo_objects(struct kmem
return x.x & OO_MASK;
}
+static inline bool cmpxchg_double_slab(struct kmem_cache *s, struct page *page,
+ void *freelist_old, unsigned long counters_old,
+ void *freelist_new, unsigned long counters_new,
+ const char *n)
+{
+#ifdef CONFIG_CMPXCHG_DOUBLE
+ if (s->flags & __CMPXCHG_DOUBLE) {
+ if (cmpxchg_double(&page->freelist,
+ freelist_old, counters_old,
+ freelist_new, counters_new))
+ return 1;
+ } else
+#endif
+ {
+ if (page->freelist == freelist_old && page->counters == counters_old) {
+ page->freelist = freelist_new;
+ page->counters = counters_new;
+ return 1;
+ }
+ }
+
+ cpu_relax();
+ stat(s, CMPXCHG_DOUBLE_FAIL);
+
+#ifdef SLUB_DEBUG_CMPXCHG
+ printk(KERN_INFO "%s %s: cmpxchg double redo ", n, s->name);
+#endif
+
+ return 0;
+}
+
/*
* Determine a map of object in use on a page.
*
@@ -2654,6 +2689,12 @@ static int kmem_cache_open(struct kmem_c
}
}
+#ifdef CONFIG_CMPXCHG_DOUBLE
+ if (system_has_cmpxchg_double() && (s->flags & SLAB_DEBUG_FLAGS) == 0)
+ /* Enable fast mode */
+ s->flags |= __CMPXCHG_DOUBLE;
+#endif
+
/*
* The larger the object size is, the more pages we want on the partial
* list to avoid pounding the page allocator excessively.
@@ -4551,6 +4592,8 @@ STAT_ATTR(DEACTIVATE_TO_HEAD, deactivate
STAT_ATTR(DEACTIVATE_TO_TAIL, deactivate_to_tail);
STAT_ATTR(DEACTIVATE_REMOTE_FREES, deactivate_remote_frees);
STAT_ATTR(ORDER_FALLBACK, order_fallback);
+STAT_ATTR(CMPXCHG_DOUBLE_CPU_FAIL, cmpxchg_double_cpu_fail);
+STAT_ATTR(CMPXCHG_DOUBLE_FAIL, cmpxchg_double_fail);
#endif
static struct attribute *slab_attrs[] = {
@@ -4608,6 +4651,8 @@ static struct attribute *slab_attrs[] =
&deactivate_to_tail_attr.attr,
&deactivate_remote_frees_attr.attr,
&order_fallback_attr.attr,
+ &cmpxchg_double_fail_attr.attr,
+ &cmpxchg_double_cpu_fail_attr.attr,
#endif
#ifdef CONFIG_FAILSLAB
&failslab_attr.attr,
Index: linux-2.6/include/linux/slub_def.h
===================================================================
--- linux-2.6.orig/include/linux/slub_def.h 2011-04-13 15:19:53.000000000 -0500
+++ linux-2.6/include/linux/slub_def.h 2011-04-15 13:14:51.000000000 -0500
@@ -33,6 +33,7 @@ enum stat_item {
DEACTIVATE_REMOTE_FREES,/* Slab contained remotely freed objects */
ORDER_FALLBACK, /* Number of times fallback was necessary */
CMPXCHG_DOUBLE_CPU_FAIL,/* Failure of this_cpu_cmpxchg_double */
+ CMPXCHG_DOUBLE_FAIL, /* Number of times that cmpxchg double did not match */
NR_SLUB_STAT_ITEMS };
struct kmem_cache_cpu {
next prev parent reply other threads:[~2011-04-15 20:48 UTC|newest]
Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-04-15 20:47 [slubllv3 00/21] SLUB: Lockless freelists for objects V3 Christoph Lameter
2011-04-15 20:47 ` [slubllv3 01/21] slub: Use NUMA_NO_NODE in get_partial Christoph Lameter
2011-04-15 20:47 ` [slubllv3 02/21] slub: get_map() function to establish map of free objects in a slab Christoph Lameter
2011-04-15 20:47 ` [slubllv3 03/21] slub: Eliminate repeated use of c->page through a new page variable Christoph Lameter
2011-04-15 20:47 ` [slubllv3 04/21] slub: Move node determination out of hotpath Christoph Lameter
2011-04-15 20:47 ` [slubllv3 05/21] slub: Move debug handlign in __slab_free Christoph Lameter
2011-04-15 20:47 ` [slubllv3 06/21] slub: Per object NUMA support Christoph Lameter
2011-04-15 20:47 ` [slubllv3 07/21] slub: Do not use frozen page flag but a bit in the page counters Christoph Lameter
2011-04-15 20:47 ` [slubllv3 08/21] slub: Move page->frozen handling near where the page->freelist handling occurs Christoph Lameter
2011-04-15 20:47 ` [slubllv3 09/21] x86: Add support for cmpxchg_double Christoph Lameter
2011-04-15 20:47 ` [slubllv3 10/21] mm: Rearrange struct page Christoph Lameter
2011-04-15 20:47 ` Christoph Lameter [this message]
2011-04-15 20:47 ` [slubllv3 12/21] slub: explicit list_lock taking Christoph Lameter
2011-04-15 20:47 ` [slubllv3 13/21] slub: Pass kmem_cache struct to lock and freeze slab Christoph Lameter
2011-04-15 20:47 ` [slubllv3 14/21] slub: Rework allocator fastpaths Christoph Lameter
2011-04-15 20:47 ` [slubllv3 15/21] slub: Invert locking and avoid slab lock Christoph Lameter
2011-04-15 20:47 ` [slubllv3 16/21] slub: Disable interrupts in free_debug processing Christoph Lameter
2011-04-15 20:47 ` [slubllv3 17/21] slub: Avoid disabling interrupts in free slowpath Christoph Lameter
2011-04-15 20:47 ` [slubllv3 18/21] slub: Get rid of the another_slab label Christoph Lameter
2011-04-15 20:47 ` [slubllv3 19/21] slub: fast release on full slab Christoph Lameter
2011-04-15 20:47 ` [slubllv3 20/21] slub: Not necessary to check for empty slab on load_freelist Christoph Lameter
2011-04-15 20:47 ` [slubllv3 21/21] slub: update statistics for cmpxchg handling Christoph Lameter
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=20110415204752.445029138@linux.com \
--to=cl@linux.com \
--cc=penberg@cs.helsinki.fi \
--cc=rientjes@google.com \
/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.