From: Pekka Enberg <penberg@cs.helsinki.fi>
To: Christoph Lameter <cl@linux-foundation.org>
Cc: Mel Gorman <mel@csn.ul.ie>, David Rientjes <rientjes@google.com>,
Tejun Heo <tj@kernel.org>,
linux-kernel@vger.kernel.org,
Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>,
Zhang Yanmin <yanmin_zhang@linux.intel.com>
Subject: Re: [this_cpu_xx V6 7/7] this_cpu: slub aggressive use of this_cpu operations in the hotpaths
Date: Wed, 14 Oct 2009 19:14:26 +0300 [thread overview]
Message-ID: <84144f020910140914u2c08655bjbd870019689ac947@mail.gmail.com> (raw)
In-Reply-To: <alpine.DEB.1.10.0910141154320.26476@gentwo.org>
Hi Christoph,
On Wed, 14 Oct 2009, Pekka Enberg wrote:
>> SLAB is able to queue lots of large objects but SLUB can't do that because it
>> has no queues. In SLUB, each CPU gets a page assigned to it that serves as a
>> "queue" but the size of the queue gets smaller as object size approaches page
>> size.
>>
>> We try to offset that with higher order allocations but IIRC we don't increase
>> the order linearly with object size and cap it to some reasonable maximum.
On Wed, Oct 14, 2009 at 6:56 PM, Christoph Lameter
<cl@linux-foundation.org> wrote:
> You can test to see if larger pages have an influence by passing
>
> slub_max_order=6
>
> or so on the kernel command line.
>
> You can force a large page use in slub by setting
>
> slub_min_order=3
>
> f.e.
>
> Or you can force a mininum number of objecxcts in slub through f.e.
>
> slub_min_objects=50
>
> slub_max_order=6 slub_min_objects=50
>
> should result in pretty large slabs with lots of in page objects that
> allow slub to queue better.
Yeah, that should help but it's probably not something we can do for
mainline. I'm not sure how we can fix SLUB to support large objects
out-of-the-box as efficiently as SLAB does.
Pekka
next prev parent reply other threads:[~2009-10-14 16:15 UTC|newest]
Thread overview: 56+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-10-07 21:10 [this_cpu_xx V6 0/7] Introduce per cpu atomic operations and avoid per cpu address arithmetic cl
2009-10-07 21:10 ` [this_cpu_xx V6 1/7] this_cpu_ops: page allocator conversion cl
2009-10-08 10:38 ` Tejun Heo
2009-10-08 10:40 ` Tejun Heo
2009-10-08 16:15 ` Christoph Lameter
2009-10-08 10:53 ` Mel Gorman
2009-10-07 21:10 ` [this_cpu_xx V6 2/7] this_cpu ops: Remove pageset_notifier cl
2009-10-07 21:10 ` [this_cpu_xx V6 3/7] Use this_cpu operations in slub cl
2009-10-12 10:19 ` Tejun Heo
2009-10-12 10:21 ` Tejun Heo
2009-10-12 14:54 ` Christoph Lameter
2009-10-13 2:13 ` Tejun Heo
2009-10-13 14:41 ` Christoph Lameter
2009-10-13 14:56 ` Tejun Heo
2009-10-13 15:20 ` Christoph Lameter
2009-10-14 1:57 ` Tejun Heo
2009-10-14 14:14 ` Christoph Lameter
2009-10-15 7:47 ` Tejun Heo
2009-10-16 16:44 ` Christoph Lameter
2009-10-18 3:11 ` Tejun Heo
2009-10-07 21:10 ` [this_cpu_xx V6 4/7] SLUB: Get rid of dynamic DMA kmalloc cache allocation cl
2009-10-13 18:48 ` [FIX] patch "SLUB: Get rid of dynamic DMA kmalloc cache allocation" Christoph Lameter
2009-10-07 21:10 ` [this_cpu_xx V6 5/7] this_cpu: Remove slub kmem_cache fields cl
2009-10-07 23:10 ` Christoph Lameter
2009-10-07 21:10 ` [this_cpu_xx V6 6/7] Make slub statistics use this_cpu_inc cl
2009-10-07 21:10 ` [this_cpu_xx V6 7/7] this_cpu: slub aggressive use of this_cpu operations in the hotpaths cl
2009-10-12 10:40 ` Tejun Heo
2009-10-12 13:14 ` Pekka Enberg
2009-10-12 14:55 ` Christoph Lameter
2009-10-13 9:45 ` David Rientjes
2009-10-13 14:43 ` Christoph Lameter
2009-10-13 19:14 ` Christoph Lameter
2009-10-13 19:44 ` Pekka Enberg
2009-10-13 19:48 ` Christoph Lameter
2009-10-13 20:15 ` David Rientjes
2009-10-13 20:28 ` Christoph Lameter
2009-10-13 22:53 ` David Rientjes
2009-10-14 13:34 ` Mel Gorman
2009-10-14 14:08 ` Christoph Lameter
2009-10-14 15:49 ` Mel Gorman
2009-10-14 15:53 ` Pekka Enberg
2009-10-14 15:56 ` Christoph Lameter
2009-10-14 16:14 ` Pekka Enberg [this message]
2009-10-14 18:19 ` Christoph Lameter
2009-10-16 10:50 ` Mel Gorman
2009-10-16 18:40 ` David Rientjes
2009-10-15 9:03 ` David Rientjes
2009-10-16 16:45 ` Christoph Lameter
2009-10-16 18:43 ` David Rientjes
2009-10-16 18:50 ` Christoph Lameter
2009-10-13 20:25 ` Christoph Lameter
2009-10-14 1:33 ` David Rientjes
2009-10-13 15:40 ` [this_cpu_xx V6 0/7] Introduce per cpu atomic operations and avoid per cpu address arithmetic Mel Gorman
2009-10-13 15:45 ` Christoph Lameter
2009-10-13 16:09 ` Mel Gorman
2009-10-13 17:17 ` 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=84144f020910140914u2c08655bjbd870019689ac947@mail.gmail.com \
--to=penberg@cs.helsinki.fi \
--cc=cl@linux-foundation.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mathieu.desnoyers@polymtl.ca \
--cc=mel@csn.ul.ie \
--cc=rientjes@google.com \
--cc=tj@kernel.org \
--cc=yanmin_zhang@linux.intel.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
all inboxes | Powered by JetHome®