From: Mel Gorman <mel@csn.ul.ie>
To: Christoph Lameter <cl@linux-foundation.org>
Cc: Linux Memory Management List <linux-mm@kvack.org>,
KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>,
Nick Piggin <npiggin@suse.de>,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
Lin Ming <ming.m.lin@intel.com>,
Zhang Yanmin <yanmin_zhang@linux.intel.com>,
Peter Zijlstra <peterz@infradead.org>,
Andrew Morton <akpm@linux-foundation.org>
Subject: Re: [PATCH 00/25] Cleanup and optimise the page allocator V5
Date: Fri, 20 Mar 2009 15:37:23 +0000 [thread overview]
Message-ID: <20090320153723.GO24586@csn.ul.ie> (raw)
In-Reply-To: <alpine.DEB.1.10.0903201059240.3740@qirst.com>
On Fri, Mar 20, 2009 at 11:00:42AM -0400, Christoph Lameter wrote:
> On Fri, 20 Mar 2009, Mel Gorman wrote:
>
> > The lock contention on some machines goes up for the the zone->lru_lock
> > and zone->lock locks which can regress some workloads even though others on
> > the same machine still go faster. For netperf, a lock called slock-AF_INET
> > seemed very important although I didn't look too closely other than noting
> > contention went up. The zone->lock gets hammered a lot by high order allocs
> > and frees coming from SLUB which are not covered by the PCP allocator in
> > this patchset. zone->lru_lock goes up is less clear but as it's page cache
> > releases but overall contention may be up because CPUs are spending less
> > time with interrupts disabled and more time trying to do real work but
> > contending on the locks.
>
> We can tune SLUB to buffer more pages if the lru lock becomes too hot.
>
hmm, I'm missing something in your reasoning. The contention I saw for
zone->lru_lock
&zone->lru_lock 37350 [<ffffffff8029d6fe>] ____pagevec_lru_add+0x9c/0x172
&zone->lru_lock 55423 [<ffffffff8029d377>] release_pages+0x10a/0x21b
&zone->lru_lock 402 [<ffffffff8029d9d9>] activate_page+0x4f/0x147
&zone->lru_lock 6 [<ffffffff8029dbbd>] put_page+0x94/0x122
So I just assumed it was LRU pages being taken off and freed that was
causing the contention. Can SLUB affect that?
Maybe you meant zone->lock and SLUB could tune buffers more to avoid
that if that lock was hot. That is one alternative but the later patches
proposed an alternative whereby high-order and compound pages could be
stored on the PCP lists. Compound only really helps SLUB but high-order
also helped stacks, signal handlers and the like so it seemed like a
good idea one way or the other. Course, this meant a search of the PCP
lists or increasing the size of the PCP structure - swings and
roundabouts :/
--
Mel Gorman
Part-time Phd Student Linux Technology Center
University of Limerick IBM Dublin Software Lab
--
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:[~2009-03-20 15:37 UTC|newest]
Thread overview: 44+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-03-20 10:02 [PATCH 00/25] Cleanup and optimise the page allocator V5 Mel Gorman
2009-03-20 10:02 ` [PATCH 01/25] Replace __alloc_pages_internal() with __alloc_pages_nodemask() Mel Gorman
2009-03-20 10:02 ` [PATCH 02/25] Do not sanity check order in the fast path Mel Gorman
2009-03-20 10:02 ` [PATCH 03/25] Do not check NUMA node ID when the caller knows the node is valid Mel Gorman
2009-03-20 10:02 ` [PATCH 04/25] Check only once if the zonelist is suitable for the allocation Mel Gorman
2009-03-20 10:02 ` [PATCH 05/25] Break up the allocator entry point into fast and slow paths Mel Gorman
2009-03-20 10:02 ` [PATCH 06/25] Move check for disabled anti-fragmentation out of fastpath Mel Gorman
2009-03-20 10:02 ` [PATCH 07/25] Check in advance if the zonelist needs additional filtering Mel Gorman
2009-03-20 15:04 ` Christoph Lameter
2009-03-20 10:02 ` [PATCH 08/25] Calculate the preferred zone for allocation only once Mel Gorman
2009-03-20 15:06 ` Christoph Lameter
2009-03-20 15:29 ` Mel Gorman
2009-03-20 10:02 ` [PATCH 09/25] Calculate the migratetype " Mel Gorman
2009-03-20 15:08 ` Christoph Lameter
2009-03-20 10:02 ` [PATCH 10/25] Calculate the alloc_flags " Mel Gorman
2009-03-20 10:02 ` [PATCH 11/25] Calculate the cold parameter " Mel Gorman
2009-03-20 15:09 ` Christoph Lameter
2009-04-21 15:13 ` Mel Gorman
2009-04-21 15:25 ` Christoph Lameter
2009-04-21 15:47 ` Mel Gorman
2009-03-20 10:02 ` [PATCH 12/25] Remove a branch by assuming __GFP_HIGH == ALLOC_HIGH Mel Gorman
2009-03-20 10:03 ` [PATCH 13/25] Inline __rmqueue_smallest() Mel Gorman
2009-03-20 10:03 ` [PATCH 14/25] Inline buffered_rmqueue() Mel Gorman
2009-03-20 10:03 ` [PATCH 15/25] Inline __rmqueue_fallback() Mel Gorman
2009-03-20 10:03 ` [PATCH 16/25] Save text by reducing call sites of __rmqueue() Mel Gorman
2009-03-20 10:03 ` [PATCH 17/25] Do not call get_pageblock_migratetype() more than necessary Mel Gorman
2009-03-20 10:03 ` [PATCH 18/25] Do not disable interrupts in free_page_mlock() Mel Gorman
2009-03-20 10:03 ` [PATCH 19/25] Do not setup zonelist cache when there is only one node Mel Gorman
2009-03-20 10:03 ` [PATCH 20/25] Do not check for compound pages during the page allocator sanity checks Mel Gorman
2009-03-20 10:03 ` [PATCH 21/25] Use allocation flags as an index to the zone watermark Mel Gorman
2009-03-20 10:03 ` [PATCH 22/25] Update NR_FREE_PAGES only as necessary Mel Gorman
2009-03-20 10:03 ` [PATCH 23/25] Get the pageblock migratetype without disabling interrupts Mel Gorman
2009-03-20 10:03 ` [PATCH 24/25] Re-sort GFP flags and fix whitespace alignment for easier reading Mel Gorman
2009-03-20 10:03 ` [PATCH 25/25] Use a pre-calculated value instead of num_online_nodes() in fast paths Mel Gorman
2009-03-20 15:00 ` [PATCH 00/25] Cleanup and optimise the page allocator V5 Christoph Lameter
2009-03-20 15:37 ` Mel Gorman [this message]
2009-03-20 16:04 ` Christoph Lameter
2009-03-20 16:41 ` Mel Gorman
2009-03-20 16:07 ` Christoph Lameter
2009-03-20 16:27 ` Mel Gorman
2009-03-20 19:43 ` Christoph Lameter
2009-03-23 11:52 ` Mel Gorman
2009-03-23 13:30 ` Christoph Lameter
2009-03-23 14:59 ` Mel Gorman
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=20090320153723.GO24586@csn.ul.ie \
--to=mel@csn.ul.ie \
--cc=akpm@linux-foundation.org \
--cc=cl@linux-foundation.org \
--cc=kosaki.motohiro@jp.fujitsu.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=ming.m.lin@intel.com \
--cc=npiggin@suse.de \
--cc=peterz@infradead.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;
as well as URLs for NNTP newsgroup(s).