From: Marek Szyprowski <m.szyprowski@samsung.com>
To: Joonsoo Kim <iamjoonsoo.kim@lge.com>,
Andrew Morton <akpm@linux-foundation.org>
Cc: Rik van Riel <riel@redhat.com>,
Johannes Weiner <hannes@cmpxchg.org>,
Mel Gorman <mgorman@suse.de>,
Laura Abbott <lauraa@codeaurora.org>,
Minchan Kim <minchan@kernel.org>,
Heesub Shin <heesub.shin@samsung.com>,
Michal Nazarewicz <mina86@mina86.com>,
linux-mm@kvack.org, linux-kernel@vger.kernel.org,
Kyungmin Park <kyungmin.park@samsung.com>,
Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>,
'Tomasz Stanislawski' <t.stanislaws@samsung.com>
Subject: Re: [RFC PATCH 0/3] Aggressively allocate the pages on cma reserved memory
Date: Fri, 09 May 2014 14:39:20 +0200 [thread overview]
Message-ID: <536CCC78.6050806@samsung.com> (raw)
In-Reply-To: <1399509144-8898-1-git-send-email-iamjoonsoo.kim@lge.com>
Hello,
On 2014-05-08 02:32, Joonsoo Kim wrote:
> This series tries to improve CMA.
>
> CMA is introduced to provide physically contiguous pages at runtime
> without reserving memory area. But, current implementation works like as
> reserving memory approach, because allocation on cma reserved region only
> occurs as fallback of migrate_movable allocation. We can allocate from it
> when there is no movable page. In that situation, kswapd would be invoked
> easily since unmovable and reclaimable allocation consider
> (free pages - free CMA pages) as free memory on the system and free memory
> may be lower than high watermark in that case. If kswapd start to reclaim
> memory, then fallback allocation doesn't occur much.
>
> In my experiment, I found that if system memory has 1024 MB memory and
> has 512 MB reserved memory for CMA, kswapd is mostly invoked around
> the 512MB free memory boundary. And invoked kswapd tries to make free
> memory until (free pages - free CMA pages) is higher than high watermark,
> so free memory on meminfo is moving around 512MB boundary consistently.
>
> To fix this problem, we should allocate the pages on cma reserved memory
> more aggressively and intelligenetly. Patch 2 implements the solution.
> Patch 1 is the simple optimization which remove useless re-trial and patch 3
> is for removing useless alloc flag, so these are not important.
> See patch 2 for more detailed description.
>
> This patchset is based on v3.15-rc4.
Thanks for posting those patches. It basically reminds me the following
discussion:
http://thread.gmane.org/gmane.linux.kernel/1391989/focus=1399524
Your approach is basically the same. I hope that your patches can be
improved
in such a way that they will be accepted by mm maintainers. I only
wonder if the
third patch is really necessary. Without it kswapd wakeup might be still
avoided
in some cases.
> Thanks.
> Joonsoo Kim (3):
> CMA: remove redundant retrying code in __alloc_contig_migrate_range
> CMA: aggressively allocate the pages on cma reserved memory when not
> used
> CMA: always treat free cma pages as non-free on watermark checking
>
> include/linux/mmzone.h | 6 +++
> mm/compaction.c | 4 --
> mm/internal.h | 3 +-
> mm/page_alloc.c | 117 +++++++++++++++++++++++++++++++++++++++---------
> 4 files changed, 102 insertions(+), 28 deletions(-)
>
Best regards
--
Marek Szyprowski, PhD
Samsung R&D Institute Poland
--
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>
WARNING: multiple messages have this Message-ID (diff)
From: Marek Szyprowski <m.szyprowski@samsung.com>
To: Joonsoo Kim <iamjoonsoo.kim@lge.com>,
Andrew Morton <akpm@linux-foundation.org>
Cc: Rik van Riel <riel@redhat.com>,
Johannes Weiner <hannes@cmpxchg.org>,
Mel Gorman <mgorman@suse.de>,
Laura Abbott <lauraa@codeaurora.org>,
Minchan Kim <minchan@kernel.org>,
Heesub Shin <heesub.shin@samsung.com>,
Michal Nazarewicz <mina86@mina86.com>,
linux-mm@kvack.org, linux-kernel@vger.kernel.org,
Kyungmin Park <kyungmin.park@samsung.com>,
Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>,
"'Tomasz Stanislawski'" <t.stanislaws@samsung.com>
Subject: Re: [RFC PATCH 0/3] Aggressively allocate the pages on cma reserved memory
Date: Fri, 09 May 2014 14:39:20 +0200 [thread overview]
Message-ID: <536CCC78.6050806@samsung.com> (raw)
In-Reply-To: <1399509144-8898-1-git-send-email-iamjoonsoo.kim@lge.com>
Hello,
On 2014-05-08 02:32, Joonsoo Kim wrote:
> This series tries to improve CMA.
>
> CMA is introduced to provide physically contiguous pages at runtime
> without reserving memory area. But, current implementation works like as
> reserving memory approach, because allocation on cma reserved region only
> occurs as fallback of migrate_movable allocation. We can allocate from it
> when there is no movable page. In that situation, kswapd would be invoked
> easily since unmovable and reclaimable allocation consider
> (free pages - free CMA pages) as free memory on the system and free memory
> may be lower than high watermark in that case. If kswapd start to reclaim
> memory, then fallback allocation doesn't occur much.
>
> In my experiment, I found that if system memory has 1024 MB memory and
> has 512 MB reserved memory for CMA, kswapd is mostly invoked around
> the 512MB free memory boundary. And invoked kswapd tries to make free
> memory until (free pages - free CMA pages) is higher than high watermark,
> so free memory on meminfo is moving around 512MB boundary consistently.
>
> To fix this problem, we should allocate the pages on cma reserved memory
> more aggressively and intelligenetly. Patch 2 implements the solution.
> Patch 1 is the simple optimization which remove useless re-trial and patch 3
> is for removing useless alloc flag, so these are not important.
> See patch 2 for more detailed description.
>
> This patchset is based on v3.15-rc4.
Thanks for posting those patches. It basically reminds me the following
discussion:
http://thread.gmane.org/gmane.linux.kernel/1391989/focus=1399524
Your approach is basically the same. I hope that your patches can be
improved
in such a way that they will be accepted by mm maintainers. I only
wonder if the
third patch is really necessary. Without it kswapd wakeup might be still
avoided
in some cases.
> Thanks.
> Joonsoo Kim (3):
> CMA: remove redundant retrying code in __alloc_contig_migrate_range
> CMA: aggressively allocate the pages on cma reserved memory when not
> used
> CMA: always treat free cma pages as non-free on watermark checking
>
> include/linux/mmzone.h | 6 +++
> mm/compaction.c | 4 --
> mm/internal.h | 3 +-
> mm/page_alloc.c | 117 +++++++++++++++++++++++++++++++++++++++---------
> 4 files changed, 102 insertions(+), 28 deletions(-)
>
Best regards
--
Marek Szyprowski, PhD
Samsung R&D Institute Poland
next prev parent reply other threads:[~2014-05-09 12:39 UTC|newest]
Thread overview: 87+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-05-08 0:32 [RFC PATCH 0/3] Aggressively allocate the pages on cma reserved memory Joonsoo Kim
2014-05-08 0:32 ` Joonsoo Kim
2014-05-08 0:32 ` [RFC PATCH 1/3] CMA: remove redundant retrying code in __alloc_contig_migrate_range Joonsoo Kim
2014-05-08 0:32 ` Joonsoo Kim
2014-05-09 15:44 ` Michal Nazarewicz
2014-05-09 15:44 ` Michal Nazarewicz
2014-05-08 0:32 ` [RFC PATCH 2/3] CMA: aggressively allocate the pages on cma reserved memory when not used Joonsoo Kim
2014-05-08 0:32 ` Joonsoo Kim
2014-05-09 15:45 ` Michal Nazarewicz
2014-05-09 15:45 ` Michal Nazarewicz
2014-05-12 17:04 ` Laura Abbott
2014-05-12 17:04 ` Laura Abbott
2014-05-13 1:14 ` Joonsoo Kim
2014-05-13 1:14 ` Joonsoo Kim
2014-05-13 3:05 ` Minchan Kim
2014-05-13 3:05 ` Minchan Kim
2014-05-24 0:57 ` Laura Abbott
2014-05-24 0:57 ` Laura Abbott
2014-05-26 2:44 ` Joonsoo Kim
2014-05-26 2:44 ` Joonsoo Kim
2014-05-13 3:00 ` Minchan Kim
2014-05-13 3:00 ` Minchan Kim
2014-05-15 1:53 ` Joonsoo Kim
2014-05-15 1:53 ` Joonsoo Kim
2014-05-15 2:43 ` Minchan Kim
2014-05-15 2:43 ` Minchan Kim
2014-05-19 2:11 ` Joonsoo Kim
2014-05-19 2:11 ` Joonsoo Kim
2014-05-19 2:53 ` Minchan Kim
2014-05-19 2:53 ` Minchan Kim
2014-05-19 4:50 ` Joonsoo Kim
2014-05-19 4:50 ` Joonsoo Kim
2014-05-19 23:18 ` Minchan Kim
2014-05-19 23:18 ` Minchan Kim
2014-05-20 6:33 ` Joonsoo Kim
2014-05-20 6:33 ` Joonsoo Kim
2014-05-15 2:45 ` Heesub Shin
2014-05-15 2:45 ` Heesub Shin
2014-05-15 5:06 ` Minchan Kim
2014-05-15 5:06 ` Minchan Kim
2014-05-19 23:22 ` Minchan Kim
2014-05-19 23:22 ` Minchan Kim
2014-05-16 8:02 ` [RFC][PATCH] CMA: drivers/base/Kconfig: restrict CMA size to non-zero value Gioh Kim
2014-05-16 8:02 ` Gioh Kim
2014-05-16 17:45 ` Michal Nazarewicz
2014-05-19 1:47 ` Gioh Kim
2014-05-19 1:47 ` Gioh Kim
2014-05-19 5:55 ` Joonsoo Kim
2014-05-19 5:55 ` Joonsoo Kim
2014-05-19 9:14 ` Gioh Kim
2014-05-19 9:14 ` Gioh Kim
2014-05-19 19:59 ` Michal Nazarewicz
2014-05-20 0:50 ` Gioh Kim
2014-05-20 0:50 ` Gioh Kim
2014-05-20 1:28 ` Michal Nazarewicz
2014-05-20 2:26 ` Gioh Kim
2014-05-20 2:26 ` Gioh Kim
2014-05-20 18:15 ` Michal Nazarewicz
2014-05-20 11:38 ` Marek Szyprowski
2014-05-20 11:38 ` Marek Szyprowski
2014-05-20 12:23 ` Gi-Oh Kim
2014-05-21 0:15 ` Gioh Kim
2014-05-21 0:15 ` Gioh Kim
2014-05-14 8:42 ` [RFC PATCH 2/3] CMA: aggressively allocate the pages on cma reserved memory when not used Aneesh Kumar K.V
2014-05-14 8:42 ` Aneesh Kumar K.V
2014-05-15 1:58 ` Joonsoo Kim
2014-05-15 1:58 ` Joonsoo Kim
2014-05-18 17:36 ` Aneesh Kumar K.V
2014-05-18 17:36 ` Aneesh Kumar K.V
2014-05-19 2:29 ` Joonsoo Kim
2014-05-19 2:29 ` Joonsoo Kim
2014-05-08 0:32 ` [RFC PATCH 3/3] CMA: always treat free cma pages as non-free on watermark checking Joonsoo Kim
2014-05-08 0:32 ` Joonsoo Kim
2014-05-09 15:46 ` Michal Nazarewicz
2014-05-09 15:46 ` Michal Nazarewicz
2014-05-09 12:39 ` Marek Szyprowski [this message]
2014-05-09 12:39 ` [RFC PATCH 0/3] Aggressively allocate the pages on cma reserved memory Marek Szyprowski
2014-05-13 2:26 ` Joonsoo Kim
2014-05-13 2:26 ` Joonsoo Kim
2014-05-14 9:44 ` Aneesh Kumar K.V
2014-05-14 9:44 ` Aneesh Kumar K.V
2014-05-15 2:10 ` Joonsoo Kim
2014-05-15 2:10 ` Joonsoo Kim
2014-05-15 9:47 ` Mel Gorman
2014-05-15 9:47 ` Mel Gorman
2014-05-19 2:12 ` Joonsoo Kim
2014-05-19 2:12 ` Joonsoo Kim
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=536CCC78.6050806@samsung.com \
--to=m.szyprowski@samsung.com \
--cc=akpm@linux-foundation.org \
--cc=b.zolnierkie@samsung.com \
--cc=hannes@cmpxchg.org \
--cc=heesub.shin@samsung.com \
--cc=iamjoonsoo.kim@lge.com \
--cc=kyungmin.park@samsung.com \
--cc=lauraa@codeaurora.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mgorman@suse.de \
--cc=mina86@mina86.com \
--cc=minchan@kernel.org \
--cc=riel@redhat.com \
--cc=t.stanislaws@samsung.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.