From: Pekka Enberg <penberg@cs.helsinki.fi>
To: Mel Gorman <mel@csn.ul.ie>
Cc: Lin Ming <ming.m.lin@intel.com>,
Linux Memory Management List <linux-mm@kvack.org>,
Rik van Riel <riel@redhat.com>,
KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>,
Christoph Lameter <cl@linux-foundation.org>,
Johannes Weiner <hannes@cmpxchg.org>,
Nick Piggin <npiggin@suse.de>,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
Zhang Yanmin <yanmin_zhang@linux.intel.com>,
Peter Zijlstra <peterz@infradead.org>
Subject: Re: [RFC PATCH 00/19] Cleanup and optimise the page allocator V2
Date: Thu, 26 Feb 2009 13:18:59 +0200 [thread overview]
Message-ID: <1235647139.16552.34.camel@penberg-laptop> (raw)
In-Reply-To: <20090226110336.GC32756@csn.ul.ie>
On Thu, 2009-02-26 at 11:03 +0000, Mel Gorman wrote:
> On Thu, Feb 26, 2009 at 05:10:27PM +0800, Lin Ming wrote:
> > We tested this v2 patch series with 2.6.29-rc6 on different machines.
> >
>
> Wonderful, thanks.
>
> > 4P qual-core 2P qual-core 2P qual-core HT
> > tigerton stockley Nehalem
> > ------------------------------------------------
> > tbench +3% +2% 0%
>
> Nice.
>
> > oltp -2% 0% 0%
>
> This is a big disappointment and somewhat confusing that it is so
> severe. For sysbench I was seeing on six different machines;
>
> 50834.14 51763.08 1.79%
> 61852.08 61966.58 0.18%
> 5935.98 5980.06 0.74%
> 29227.78 30167.72 3.12%
> 66702.67 66534.76 -0.25%
> 26643.18 26542.59 -0.38%
>
> So, two smallish regressions but mainly gains. Then again, I'm becoming
> more and more convinced that sysbench doesn't really represent a proper
> OLTP workload.
>
> I'd like to understand more how the page allocator at least was being used
> during your tests. Would it be possible to get a full profile (including
> instruction if possible and the vmlinux file) for both kernels please?
>
> If you can get the profiles, confirm the regression is still there as
> sometimes profiling can alter the outcome. Even if this happens, the
> profile will tell me where time is being spent.
>
> > aim7 0% 0% 0%
> > specjbb2005 +3% 0% 0%
> > hackbench 0% 0% 0%
> >
> > netperf:
> > TCP-S-112k 0% -1% 0%
> > TCP-S-64k 0% -1% +1%
> > TCP-RR-1 0% 0% +1%
> > UDP-U-4k -2% 0% -2%
>
> Pekka, for this test was SLUB or the page allocator handling the 4K
> allocations?
The page allocator. The pass-through revert is not in 2.6.29-rc6 and I
won't be sending it until 2.6.30 opens up.
>
> > UDP-U-1k +3% 0% 0%
> > UDP-RR-1 0% 0% 0%
> > UDP-RR-512 -1% 0% +1%
> >
> > Lin Ming
> >
>
> Thanks a million for testing.
>
--
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:[~2009-02-26 11:19 UTC|newest]
Thread overview: 59+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-02-24 12:16 Mel Gorman
2009-02-24 12:16 ` [PATCH 01/19] Replace __alloc_pages_internal() with __alloc_pages_nodemask() Mel Gorman
2009-02-24 12:16 ` [PATCH 02/19] Do not sanity check order in the fast path Mel Gorman
2009-02-24 12:16 ` [PATCH 03/19] Do not check NUMA node ID when the caller knows the node is valid Mel Gorman
2009-02-24 17:17 ` Christoph Lameter
2009-02-24 12:17 ` [PATCH 04/19] Convert gfp_zone() to use a table of precalculated values Mel Gorman
2009-02-24 16:43 ` Christoph Lameter
2009-02-24 17:07 ` Mel Gorman
2009-02-24 12:17 ` [PATCH 05/19] Re-sort GFP flags and fix whitespace alignment for easier reading Mel Gorman
2009-02-24 12:17 ` [PATCH 06/19] Check only once if the zonelist is suitable for the allocation Mel Gorman
2009-02-24 17:24 ` Christoph Lameter
2009-02-24 12:17 ` [PATCH 07/19] Break up the allocator entry point into fast and slow paths Mel Gorman
2009-02-24 12:17 ` [PATCH 08/19] Simplify the check on whether cpusets are a factor or not Mel Gorman
2009-02-24 17:27 ` Christoph Lameter
2009-02-24 17:55 ` Mel Gorman
2009-02-24 12:17 ` [PATCH 09/19] Move check for disabled anti-fragmentation out of fastpath Mel Gorman
2009-02-24 12:17 ` [PATCH 10/19] Calculate the preferred zone for allocation only once Mel Gorman
2009-02-24 17:31 ` Christoph Lameter
2009-02-24 17:53 ` Mel Gorman
2009-02-24 12:17 ` [PATCH 11/19] Calculate the migratetype " Mel Gorman
2009-02-24 12:17 ` [PATCH 12/19] Calculate the alloc_flags " Mel Gorman
2009-02-24 12:17 ` [PATCH 13/19] Inline __rmqueue_smallest() Mel Gorman
2009-02-24 12:17 ` [PATCH 14/19] Inline buffered_rmqueue() Mel Gorman
2009-02-24 12:17 ` [PATCH 15/19] Do not call get_pageblock_migratetype() more than necessary Mel Gorman
2009-02-24 12:17 ` [PATCH 16/19] Do not disable interrupts in free_page_mlock() Mel Gorman
2009-02-24 12:17 ` [PATCH 17/19] Do not setup zonelist cache when there is only one node Mel Gorman
2009-02-24 12:17 ` [PATCH 18/19] Do not check for compound pages during the page allocator sanity checks Mel Gorman
2009-02-24 12:17 ` [PATCH 19/19] Split per-cpu list into one-list-per-migrate-type Mel Gorman
2009-02-26 9:10 ` [RFC PATCH 00/19] Cleanup and optimise the page allocator V2 Lin Ming
2009-02-26 9:26 ` Pekka Enberg
2009-02-26 9:27 ` Lin Ming
2009-02-26 11:03 ` Mel Gorman
2009-02-26 11:18 ` Pekka Enberg [this message]
2009-02-26 11:22 ` Mel Gorman
2009-02-26 12:27 ` Lin Ming
2009-02-27 8:44 ` Lin Ming
2009-03-02 11:21 ` Mel Gorman
2009-03-02 11:39 ` Nick Piggin
2009-03-02 12:16 ` Mel Gorman
2009-03-03 4:42 ` Nick Piggin
2009-03-03 8:25 ` Mel Gorman
2009-03-03 9:04 ` Nick Piggin
2009-03-03 13:51 ` Mel Gorman
2009-03-03 16:31 ` Christoph Lameter
2009-03-03 21:48 ` Mel Gorman
2009-03-04 2:05 ` Zhang, Yanmin
2009-03-04 7:23 ` Peter Zijlstra
2009-03-04 8:31 ` Zhang, Yanmin
2009-03-04 9:07 ` Nick Piggin
2009-03-05 1:56 ` Zhang, Yanmin
2009-03-05 10:34 ` Ingo Molnar
2009-03-06 8:33 ` Lin Ming
2009-03-06 9:39 ` Ingo Molnar
2009-03-06 13:03 ` Mel Gorman
2009-03-09 1:50 ` Zhang, Yanmin
2009-03-09 7:31 ` Lin Ming
2009-03-09 7:03 ` Lin Ming
2009-03-04 18:04 ` Mel Gorman
2009-02-26 16:28 ` 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=1235647139.16552.34.camel@penberg-laptop \
--to=penberg@cs.helsinki.fi \
--cc=cl@linux-foundation.org \
--cc=hannes@cmpxchg.org \
--cc=kosaki.motohiro@jp.fujitsu.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mel@csn.ul.ie \
--cc=ming.m.lin@intel.com \
--cc=npiggin@suse.de \
--cc=peterz@infradead.org \
--cc=riel@redhat.com \
--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