From: Michal Hocko <mhocko@suse.cz>
To: Glauber Costa <glommer@parallels.com>
Cc: linux-kernel@vger.kernel.org, linux-mm@kvack.org,
cgroups@vger.kernel.org, devel@openvz.org,
Johannes Weiner <hannes@cmpxchg.org>,
Andrew Morton <akpm@linux-foundation.org>,
kamezawa.hiroyu@jp.fujitsu.com, Christoph Lameter <cl@linux.com>,
David Rientjes <rientjes@google.com>,
Pekka Enberg <penberg@kernel.org>,
Suleiman Souhlal <suleiman@google.com>
Subject: Re: [PATCH v2 02/11] memcg: Reclaim when more than one page needed.
Date: Fri, 10 Aug 2012 20:54:17 +0200 [thread overview]
Message-ID: <20120810185417.GB16110@dhcp22.suse.cz> (raw)
In-Reply-To: <1344517279-30646-3-git-send-email-glommer@parallels.com>
On Thu 09-08-12 17:01:10, Glauber Costa wrote:
> From: Suleiman Souhlal <ssouhlal@FreeBSD.org>
>
> mem_cgroup_do_charge() was written before kmem accounting, and expects
> three cases: being called for 1 page, being called for a stock of 32
> pages, or being called for a hugepage. If we call for 2 or 3 pages (and
> both the stack and several slabs used in process creation are such, at
> least with the debug options I had), it assumed it's being called for
> stock and just retried without reclaiming.
>
> Fix that by passing down a minsize argument in addition to the csize.
>
> And what to do about that (csize == PAGE_SIZE && ret) retry? If it's
> needed at all (and presumably is since it's there, perhaps to handle
> races), then it should be extended to more than PAGE_SIZE, yet how far?
> And should there be a retry count limit, of what? For now retry up to
> COSTLY_ORDER (as page_alloc.c does) and make sure not to do it if
> __GFP_NORETRY.
>
> [v4: fixed nr pages calculation pointed out by Christoph Lameter ]
>
> Signed-off-by: Suleiman Souhlal <suleiman@google.com>
> Signed-off-by: Glauber Costa <glommer@parallels.com>
> Reviewed-by: Kamezawa Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
I am not happy with the min_pages argument but we can do something more
clever later.
Acked-by: Michal Hocko <mhocko@suse.cz>
> ---
> mm/memcontrol.c | 16 +++++++++-------
> 1 file changed, 9 insertions(+), 7 deletions(-)
>
> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> index bc7bfa7..2cef99a 100644
> --- a/mm/memcontrol.c
> +++ b/mm/memcontrol.c
> @@ -2294,7 +2294,8 @@ enum {
> };
>
> static int mem_cgroup_do_charge(struct mem_cgroup *memcg, gfp_t gfp_mask,
> - unsigned int nr_pages, bool oom_check)
> + unsigned int nr_pages, unsigned int min_pages,
> + bool oom_check)
> {
> unsigned long csize = nr_pages * PAGE_SIZE;
> struct mem_cgroup *mem_over_limit;
> @@ -2317,18 +2318,18 @@ static int mem_cgroup_do_charge(struct mem_cgroup *memcg, gfp_t gfp_mask,
> } else
> mem_over_limit = mem_cgroup_from_res_counter(fail_res, res);
> /*
> - * nr_pages can be either a huge page (HPAGE_PMD_NR), a batch
> - * of regular pages (CHARGE_BATCH), or a single regular page (1).
> - *
> * Never reclaim on behalf of optional batching, retry with a
> * single page instead.
> */
> - if (nr_pages == CHARGE_BATCH)
> + if (nr_pages > min_pages)
> return CHARGE_RETRY;
>
> if (!(gfp_mask & __GFP_WAIT))
> return CHARGE_WOULDBLOCK;
>
> + if (gfp_mask & __GFP_NORETRY)
> + return CHARGE_NOMEM;
> +
> ret = mem_cgroup_reclaim(mem_over_limit, gfp_mask, flags);
> if (mem_cgroup_margin(mem_over_limit) >= nr_pages)
> return CHARGE_RETRY;
> @@ -2341,7 +2342,7 @@ static int mem_cgroup_do_charge(struct mem_cgroup *memcg, gfp_t gfp_mask,
> * unlikely to succeed so close to the limit, and we fall back
> * to regular pages anyway in case of failure.
> */
> - if (nr_pages == 1 && ret)
> + if (nr_pages <= (1 << PAGE_ALLOC_COSTLY_ORDER) && ret)
> return CHARGE_RETRY;
>
> /*
> @@ -2476,7 +2477,8 @@ again:
> nr_oom_retries = MEM_CGROUP_RECLAIM_RETRIES;
> }
>
> - ret = mem_cgroup_do_charge(memcg, gfp_mask, batch, oom_check);
> + ret = mem_cgroup_do_charge(memcg, gfp_mask, batch, nr_pages,
> + oom_check);
> switch (ret) {
> case CHARGE_OK:
> break;
> --
> 1.7.11.2
>
> --
> To unsubscribe from this list: send the line "unsubscribe cgroups" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
--
Michal Hocko
SUSE Labs
next prev parent reply other threads:[~2012-08-10 18:54 UTC|newest]
Thread overview: 135+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-08-09 13:01 [PATCH v2 00/11] Request for Inclusion: kmem controller for memcg Glauber Costa
2012-08-09 13:01 ` [PATCH v2 01/11] memcg: Make it possible to use the stock for more than one page Glauber Costa
2012-08-10 15:12 ` Michal Hocko
2012-08-09 13:01 ` [PATCH v2 02/11] memcg: Reclaim when more than one page needed Glauber Costa
2012-08-10 15:42 ` Michal Hocko
2012-08-10 16:49 ` Kamezawa Hiroyuki
2012-08-10 17:28 ` Michal Hocko
2012-08-10 17:56 ` Kamezawa Hiroyuki
2012-08-10 17:30 ` Michal Hocko
2012-08-10 18:52 ` Michal Hocko
2012-08-10 18:54 ` Michal Hocko [this message]
2012-08-13 8:05 ` Glauber Costa
2012-08-13 13:10 ` Michal Hocko
2012-08-09 13:01 ` [PATCH v2 03/11] memcg: change defines to an enum Glauber Costa
2012-08-10 15:43 ` Michal Hocko
2012-08-09 13:01 ` [PATCH v2 04/11] kmem accounting basic infrastructure Glauber Costa
2012-08-10 17:02 ` Kamezawa Hiroyuki
2012-08-13 8:36 ` Glauber Costa
2012-08-17 2:38 ` Kamezawa Hiroyuki
2012-08-14 16:21 ` Michal Hocko
2012-08-15 9:33 ` Glauber Costa
2012-08-15 11:12 ` James Bottomley
2012-08-15 12:55 ` Michal Hocko
2012-08-15 13:29 ` James Bottomley
2012-08-15 12:39 ` Michal Hocko
2012-08-15 12:53 ` Glauber Costa
2012-08-15 13:02 ` Michal Hocko
2012-08-15 13:04 ` Glauber Costa
2012-08-15 13:26 ` Michal Hocko
2012-08-15 13:31 ` Glauber Costa
2012-08-15 14:10 ` Michal Hocko
2012-08-15 14:11 ` Glauber Costa
2012-08-15 14:47 ` Christoph Lameter
2012-08-15 15:11 ` Glauber Costa
2012-08-15 15:34 ` Christoph Lameter
2012-08-15 15:35 ` Glauber Costa
2012-08-15 17:26 ` Christoph Lameter
2012-08-15 18:11 ` Ying Han
2012-08-15 18:25 ` Christoph Lameter
2012-08-15 19:22 ` Glauber Costa
2012-08-15 18:07 ` Ying Han
2012-08-15 15:19 ` Greg Thelen
2012-08-15 15:36 ` Christoph Lameter
2012-08-15 18:01 ` Ying Han
2012-08-15 18:00 ` Glauber Costa
2012-08-15 19:50 ` Ying Han
2012-08-16 15:25 ` Michal Hocko
2012-08-17 5:58 ` Ying Han
2012-08-09 13:01 ` [PATCH v2 05/11] Add a __GFP_KMEMCG flag Glauber Costa
2012-08-10 17:07 ` Kamezawa Hiroyuki
2012-08-09 13:01 ` [PATCH v2 06/11] memcg: kmem controller infrastructure Glauber Costa
2012-08-10 17:27 ` Kamezawa Hiroyuki
2012-08-13 8:28 ` Glauber Costa
2012-08-14 18:58 ` Greg Thelen
2012-08-15 9:18 ` Glauber Costa
2012-08-15 16:38 ` Greg Thelen
2012-08-15 17:00 ` Glauber Costa
2012-08-15 17:12 ` Greg Thelen
2012-08-15 19:31 ` Glauber Costa
2012-08-16 3:37 ` Greg Thelen
2012-08-16 7:47 ` Glauber Costa
2012-08-20 13:36 ` Kamezawa Hiroyuki
2012-08-20 15:29 ` Glauber Costa
2012-08-17 2:36 ` Kamezawa Hiroyuki
2012-08-17 7:04 ` Glauber Costa
2012-08-14 11:00 ` Glauber Costa
2012-08-11 5:11 ` Greg Thelen
2012-08-13 8:07 ` Glauber Costa
2012-08-13 9:59 ` Glauber Costa
2012-08-13 21:21 ` Greg Thelen
2012-08-14 17:25 ` Michal Hocko
2012-08-15 9:42 ` Glauber Costa
2012-08-15 10:44 ` Glauber Costa
2012-08-15 13:09 ` Michal Hocko
2012-08-15 14:01 ` Glauber Costa
2012-08-15 14:23 ` Michal Hocko
2012-08-15 14:27 ` Glauber Costa
2012-08-16 9:53 ` Michal Hocko
2012-08-16 9:57 ` Glauber Costa
2012-08-16 15:05 ` Michal Hocko
2012-08-16 15:22 ` Glauber Costa
2012-08-21 21:50 ` Greg Thelen
2012-08-22 8:35 ` Glauber Costa
2012-08-23 0:07 ` Greg Thelen
2012-08-23 7:51 ` Glauber Costa
2012-08-09 13:01 ` [PATCH v2 07/11] mm: Allocate kernel pages to the right memcg Glauber Costa
2012-08-09 16:33 ` Greg Thelen
2012-08-09 16:42 ` Glauber Costa
2012-08-10 17:33 ` Kamezawa Hiroyuki
2012-08-13 8:03 ` Glauber Costa
2012-08-13 8:57 ` Mel Gorman
2012-08-10 17:36 ` Greg Thelen
2012-08-13 8:02 ` Glauber Costa
2012-08-14 15:16 ` Mel Gorman
2012-08-15 9:08 ` Glauber Costa
2012-08-15 13:22 ` Mel Gorman
2012-08-15 13:39 ` Glauber Costa
2012-08-15 13:51 ` Glauber Costa
2012-08-15 9:24 ` Michal Hocko
2012-08-09 13:01 ` [PATCH v2 08/11] memcg: disable kmem code when not in use Glauber Costa
2012-08-17 7:02 ` Michal Hocko
2012-08-17 7:01 ` Glauber Costa
2012-08-17 8:04 ` Michal Hocko
2012-08-09 13:01 ` [PATCH v2 09/11] memcg: propagate kmem limiting information to children Glauber Costa
2012-08-10 17:51 ` Kamezawa Hiroyuki
2012-08-13 8:01 ` Glauber Costa
2012-08-17 9:00 ` Michal Hocko
2012-08-17 9:15 ` Glauber Costa
2012-08-17 9:35 ` Michal Hocko
2012-08-17 10:07 ` Glauber Costa
2012-08-17 10:35 ` Michal Hocko
2012-08-17 10:36 ` Glauber Costa
2012-08-21 7:54 ` Michal Hocko
2012-08-21 8:35 ` Michal Hocko
2012-08-21 9:17 ` Glauber Costa
2012-08-21 9:22 ` Glauber Costa
2012-08-21 10:00 ` Michal Hocko
2012-08-21 10:01 ` Glauber Costa
2012-08-22 1:09 ` Greg Thelen
2012-08-22 8:22 ` Glauber Costa
2012-08-22 23:23 ` Greg Thelen
2012-08-23 7:55 ` Glauber Costa
2012-08-24 5:06 ` Greg Thelen
2012-08-24 5:23 ` Glauber Costa
2012-08-17 10:39 ` Glauber Costa
2012-08-09 13:01 ` [PATCH v2 10/11] memcg: allow a memcg with kmem charges to be destructed Glauber Costa
2012-08-21 8:22 ` Michal Hocko
2012-08-22 8:36 ` Glauber Costa
2012-08-09 13:01 ` [PATCH v2 11/11] protect architectures where THREAD_SIZE >= PAGE_SIZE against fork bombs Glauber Costa
2012-08-10 17:54 ` Kamezawa Hiroyuki
2012-08-21 9:35 ` Michal Hocko
2012-08-21 9:40 ` Glauber Costa
2012-08-21 10:57 ` Michal Hocko
2012-08-17 21:37 ` [PATCH v2 00/11] Request for Inclusion: kmem controller for memcg Ying Han
2012-08-20 7:51 ` Glauber Costa
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=20120810185417.GB16110@dhcp22.suse.cz \
--to=mhocko@suse.cz \
--cc=akpm@linux-foundation.org \
--cc=cgroups@vger.kernel.org \
--cc=cl@linux.com \
--cc=devel@openvz.org \
--cc=glommer@parallels.com \
--cc=hannes@cmpxchg.org \
--cc=kamezawa.hiroyu@jp.fujitsu.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=penberg@kernel.org \
--cc=rientjes@google.com \
--cc=suleiman@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).