From: Christoph Lameter <cl@linux.com>
To: Pekka Enberg <penberg@cs.helsinki.fi>
Cc: linux-mm@kvack.org, linux-kernel@vger.kernel.org,
David Rientjes <rientjes@google.com>
Subject: [UnifiedV4 14/16] slub: Reduce size of not performance critical slabs
Date: Tue, 05 Oct 2010 13:57:39 -0500 [thread overview]
Message-ID: <20101005185819.929054093@linux.com> (raw)
In-Reply-To: 20101005185725.088808842@linux.com
[-- Attachment #1: unified_reduce --]
[-- Type: text/plain, Size: 4026 bytes --]
There are some slab caches around that are rarely used and which are not
performance critical. Add a new SLAB_LOWMEM option to reduce the memory
requirements of such slabs. SLAB_LOWMEM caches will keep no empty slabs
around, have no shared or alien caches and will have a small per cpu
queue of 5 objects.
Signed-off-by: Christoph Lameter <cl@linux.com>
---
include/linux/slab.h | 2 ++
mm/slub.c | 25 ++++++++++++++++---------
2 files changed, 18 insertions(+), 9 deletions(-)
Index: linux-2.6/include/linux/slab.h
===================================================================
--- linux-2.6.orig/include/linux/slab.h 2010-10-05 13:40:04.000000000 -0500
+++ linux-2.6/include/linux/slab.h 2010-10-05 13:40:08.000000000 -0500
@@ -17,12 +17,14 @@
* The ones marked DEBUG are only valid if CONFIG_SLAB_DEBUG is set.
*/
#define SLAB_DEBUG_FREE 0x00000100UL /* DEBUG: Perform (expensive) checks on free */
+#define SLAB_LOWMEM 0x00000200UL /* Reduce memory usage of this slab */
#define SLAB_RED_ZONE 0x00000400UL /* DEBUG: Red zone objs in a cache */
#define SLAB_POISON 0x00000800UL /* DEBUG: Poison objects */
#define SLAB_HWCACHE_ALIGN 0x00002000UL /* Align objs on cache lines */
#define SLAB_CACHE_DMA 0x00004000UL /* Use GFP_DMA memory */
#define SLAB_STORE_USER 0x00010000UL /* DEBUG: Store the last owner for bug hunting */
#define SLAB_PANIC 0x00040000UL /* Panic if kmem_cache_create() fails */
+
/*
* SLAB_DESTROY_BY_RCU - **WARNING** READ THIS!
*
Index: linux-2.6/mm/slub.c
===================================================================
--- linux-2.6.orig/mm/slub.c 2010-10-05 13:40:04.000000000 -0500
+++ linux-2.6/mm/slub.c 2010-10-05 13:40:08.000000000 -0500
@@ -2829,12 +2829,20 @@ static int kmem_cache_open(struct kmem_c
* The larger the object size is, the more pages we want on the partial
* list to avoid pounding the page allocator excessively.
*/
- set_min_partial(s, ilog2(s->size));
+ if (flags & SLAB_LOWMEM)
+ set_min_partial(s, 0);
+ else
+ set_min_partial(s, ilog2(s->size));
+
s->refcount = 1;
if (!init_kmem_cache_nodes(s))
goto error;
- s->queue = initial_queue_size(s->size);
+ if (flags & SLAB_LOWMEM)
+ s->queue = 5;
+ else
+ s->queue = initial_queue_size(s->size);
+
s->batch = (s->queue + 1) / 2;
#ifdef CONFIG_NUMA
@@ -2879,7 +2887,9 @@ static int kmem_cache_open(struct kmem_c
if (alloc_kmem_cache_cpus(s)) {
s->shared_queue_sysfs = 0;
- if (nr_cpu_ids > 1 && s->size < PAGE_SIZE) {
+ if (!(flags & SLAB_LOWMEM) &&
+ nr_cpu_ids > 1 &&
+ s->size < PAGE_SIZE) {
s->shared_queue_sysfs = 10 * s->batch;
alloc_shared_caches(s);
}
@@ -3788,7 +3798,7 @@ void __init kmem_cache_init(void)
kmem_cache_open(kmem_cache_node, "kmem_cache_node",
sizeof(struct kmem_cache_node),
- 0, SLAB_HWCACHE_ALIGN | SLAB_PANIC, NULL);
+ 0, SLAB_HWCACHE_ALIGN | SLAB_PANIC | SLAB_LOWMEM, NULL);
hotplug_memory_notifier(slab_memory_callback, SLAB_CALLBACK_PRI);
@@ -3797,7 +3807,7 @@ void __init kmem_cache_init(void)
temp_kmem_cache = kmem_cache;
kmem_cache_open(kmem_cache, "kmem_cache", kmem_size,
- 0, SLAB_HWCACHE_ALIGN | SLAB_PANIC, NULL);
+ 0, SLAB_HWCACHE_ALIGN | SLAB_PANIC | SLAB_LOWMEM, NULL);
kmem_cache = kmem_cache_alloc(kmem_cache, GFP_NOWAIT);
memcpy(kmem_cache, temp_kmem_cache, kmem_size);
@@ -3906,12 +3916,9 @@ void __init kmem_cache_init(void)
if (s && s->size) {
char *name = kasprintf(GFP_NOWAIT,
"dma-kmalloc-%d", s->objsize);
-
BUG_ON(!name);
kmalloc_dma_caches[i] = create_kmalloc_cache(name,
- s->objsize, SLAB_CACHE_DMA);
- /* DMA caches are rarely used. Reduce memory consumption */
- kmalloc_dma_caches[i]->shared_queue_sysfs = 0;
+ s->objsize, SLAB_CACHE_DMA | SLAB_LOWMEM);
}
}
#endif
--
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/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
next prev parent reply other threads:[~2010-10-05 18:58 UTC|newest]
Thread overview: 48+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-10-05 18:57 [UnifiedV4 00/16] The Unified slab allocator (V4) Christoph Lameter
2010-10-05 18:57 ` [UnifiedV4 01/16] slub: Enable sysfs support for !CONFIG_SLUB_DEBUG Christoph Lameter
2010-10-06 14:02 ` Pekka Enberg
2010-10-05 18:57 ` [UnifiedV4 02/16] slub: Move functions to reduce #ifdefs Christoph Lameter
2010-10-06 14:02 ` Pekka Enberg
2010-10-05 18:57 ` [UnifiedV4 03/16] slub: Add per cpu queueing Christoph Lameter
2010-10-05 18:57 ` [UnifiedV4 04/16] slub: Allow resizing of per cpu queues Christoph Lameter
2010-10-05 18:57 ` [UnifiedV4 05/16] slub: Remove MAX_OBJS limitation Christoph Lameter
2010-10-05 18:57 ` [UnifiedV4 06/16] slub: Drop allocator announcement Christoph Lameter
2010-10-05 18:57 ` [UnifiedV4 07/16] slub: Object based NUMA policies Christoph Lameter
2010-10-05 18:57 ` [UnifiedV4 08/16] slub: Get rid of page lock and rely on per node lock Christoph Lameter
2010-10-05 18:57 ` [UnifiedV4 09/16] slub: Shared cache to exploit cross cpu caching abilities Christoph Lameter
2010-10-05 18:57 ` [UnifiedV4 10/16] slub: Support Alien Caches Christoph Lameter
2010-10-05 18:57 ` [UnifiedV4 11/16] slub: Add a "touched" state to queues and partial lists Christoph Lameter
2010-10-05 18:57 ` [UnifiedV4 12/16] slub: Cached object expiration Christoph Lameter
2010-10-05 18:57 ` [UnifiedV4 13/16] vmscan: Tie slub object expiration into page reclaim Christoph Lameter
2010-10-05 18:57 ` Christoph Lameter [this message]
2010-10-05 18:57 ` [UnifiedV4 15/16] slub: Detailed reports on validate Christoph Lameter
2010-10-05 18:57 ` [UnifiedV4 16/16] slub: Add stats for alien allocation slowpath Christoph Lameter
2010-10-06 8:01 ` [UnifiedV4 00/16] The Unified slab allocator (V4) Pekka Enberg
2010-10-06 11:03 ` Richard Kennedy
2010-10-06 11:19 ` Pekka Enberg
2010-10-06 15:46 ` Richard Kennedy
2010-10-06 16:21 ` [UnifiedV4 slabinfo 1/2] Move slabinfo.c to tools/slub/slabinfo.c Christoph Lameter
2010-10-06 16:21 ` [UnifiedV4 slabinfo 2/2] slub: update slabinfo.c for queuing Christoph Lameter
2010-10-06 20:56 ` [UnifiedV4 00/16] The Unified slab allocator (V4) Christoph Lameter
2010-10-06 16:00 ` Christoph Lameter
2010-10-06 12:37 ` Wu Fengguang
2010-10-13 2:21 ` Alex,Shi
2010-10-18 18:00 ` Christoph Lameter
2010-10-19 0:01 ` Alex,Shi
2010-10-06 15:56 ` Christoph Lameter
2010-10-13 14:14 ` Mel Gorman
2010-10-18 18:13 ` Christoph Lameter
2010-10-19 9:23 ` Mel Gorman
2010-10-12 18:25 ` Mel Gorman
2010-10-13 7:16 ` Pekka Enberg
2010-10-13 13:46 ` Mel Gorman
2010-10-13 16:10 ` Christoph Lameter
2010-10-06 10:47 ` Andi Kleen
2010-10-06 15:59 ` Christoph Lameter
2010-10-06 16:25 ` Andi Kleen
2010-10-06 16:37 ` Christoph Lameter
2010-10-06 16:43 ` Andi Kleen
2010-10-06 16:49 ` Christoph Lameter
2010-10-06 16:52 ` Christoph Lameter
2010-10-19 20:39 ` David Rientjes
2010-10-20 13:47 ` 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=20101005185819.929054093@linux.com \
--to=cl@linux.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).