LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
From: Vladimir Davydov <vdavydov.dev@gmail.com>
To: Kirill Tkhai <ktkhai@virtuozzo.com>
Cc: akpm@linux-foundation.org, shakeelb@google.com,
	viro@zeniv.linux.org.uk, hannes@cmpxchg.org, mhocko@kernel.org,
	tglx@linutronix.de, pombredanne@nexb.com,
	stummala@codeaurora.org, gregkh@linuxfoundation.org,
	sfr@canb.auug.org.au, guro@fb.com, mka@chromium.org,
	penguin-kernel@I-love.SAKURA.ne.jp, chris@chris-wilson.co.uk,
	longman@redhat.com, minchan@kernel.org, hillf.zj@alibaba-inc.com,
	ying.huang@intel.com, mgorman@techsingularity.net, jbacik@fb.com,
	linux@roeck-us.net, linux-kernel@vger.kernel.org,
	linux-mm@kvack.org, willy@infradead.org, lirongqing@baidu.com,
	aryabinin@virtuozzo.com
Subject: Re: [PATCH v2 12/12] mm: Clear shrinker bit if there are no objects related to memcg
Date: Sun, 22 Apr 2018 21:21:32 +0300	[thread overview]
Message-ID: <20180422182132.c4tqkyy4ojgi7l7q@esperanza> (raw)
In-Reply-To: <152399129187.3456.5685999465635300270.stgit@localhost.localdomain>

On Tue, Apr 17, 2018 at 09:54:51PM +0300, Kirill Tkhai wrote:
> To avoid further unneed calls of do_shrink_slab()
> for shrinkers, which already do not have any charged
> objects in a memcg, their bits have to be cleared.
> 
> This patch introduces a lockless mechanism to do that
> without races without parallel list lru add. After
> do_shrink_slab() returns SHRINK_EMPTY the first time,
> we clear the bit and call it once again. Then we restore
> the bit, if the new return value is different.
> 
> Note, that single smp_mb__after_atomic() in shrink_slab_memcg()
> covers two situations:
> 
> 1)list_lru_add()     shrink_slab_memcg
>     list_add_tail()    for_each_set_bit() <--- read bit
>                          do_shrink_slab() <--- missed list update (no barrier)
>     <MB>                 <MB>
>     set_bit()            do_shrink_slab() <--- seen list update
> 
> This situation, when the first do_shrink_slab() sees set bit,
> but it doesn't see list update (i.e., race with the first element
> queueing), is rare. So we don't add <MB> before the first call
> of do_shrink_slab() instead of this to do not slow down generic
> case. Also, it's need the second call as seen in below in (2).
> 
> 2)list_lru_add()      shrink_slab_memcg()
>     list_add_tail()     ...
>     set_bit()           ...
>   ...                   for_each_set_bit()
>   do_shrink_slab()        do_shrink_slab()
>     clear_bit()           ...
>   ...                     ...
>   list_lru_add()          ...
>     list_add_tail()       clear_bit()
>     <MB>                  <MB>
>     set_bit()             do_shrink_slab()
> 
> The barriers guarantees, the second do_shrink_slab()
> in the right side task sees list update if really
> cleared the bit. This case is drawn in the code comment.
> 
> [Results/performance of the patchset]
> 
> After the whole patchset applied the below test shows signify
> increase of performance:
> 
> $echo 1 > /sys/fs/cgroup/memory/memory.use_hierarchy
> $mkdir /sys/fs/cgroup/memory/ct
> $echo 4000M > /sys/fs/cgroup/memory/ct/memory.kmem.limit_in_bytes
>     $for i in `seq 0 4000`; do mkdir /sys/fs/cgroup/memory/ct/$i; echo $$ > /sys/fs/cgroup/memory/ct/$i/cgroup.procs; mkdir -p s/$i; mount -t tmpfs $i s/$i; touch s/$i/file; done
> 
> Then, 4 sequential calls of drop caches:
> $time echo 3 > /proc/sys/vm/drop_caches
> 
> 1)Before:
> 0.00user 8.99system 0:08.99elapsed 99%CPU
> 0.00user 5.97system 0:05.97elapsed 100%CPU
> 0.00user 5.97system 0:05.97elapsed 100%CPU
> 0.00user 5.85system 0:05.85elapsed 100%CPU
> 
> 2)After
> 0.00user 1.11system 0:01.12elapsed 99%CPU
> 0.00user 0.00system 0:00.00elapsed 100%CPU
> 0.00user 0.00system 0:00.00elapsed 100%CPU
> 0.00user 0.00system 0:00.00elapsed 100%CPU
> 
> Even if we round 0:00.00 up to 0:00.01, the results shows
> the performance increases at least in 585 times.
> 
> Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
> ---
>  include/linux/memcontrol.h |    2 ++
>  mm/vmscan.c                |   19 +++++++++++++++++--
>  2 files changed, 19 insertions(+), 2 deletions(-)
> 
> diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
> index e1c1fa8e417a..1c5c68550e2f 100644
> --- a/include/linux/memcontrol.h
> +++ b/include/linux/memcontrol.h
> @@ -1245,6 +1245,8 @@ static inline void set_shrinker_bit(struct mem_cgroup *memcg, int nid, int nr)
>  
>  		rcu_read_lock();
>  		map = SHRINKERS_MAP(memcg, nid);
> +		/* Pairs with smp mb in shrink_slab() */
> +		smp_mb__before_atomic();
>  		set_bit(nr, map->map);
>  		rcu_read_unlock();
>  	}
> diff --git a/mm/vmscan.c b/mm/vmscan.c
> index 3be9b4d81c13..a8733bc5377b 100644
> --- a/mm/vmscan.c
> +++ b/mm/vmscan.c
> @@ -579,8 +579,23 @@ static unsigned long shrink_slab_memcg(gfp_t gfp_mask, int nid,
>  		}
>  
>  		ret = do_shrink_slab(&sc, shrinker, priority);
> -		if (ret == SHRINK_EMPTY)
> -			ret = 0;
> +		if (ret == SHRINK_EMPTY) {
> +			clear_bit(i, map->map);
> +			/*
> +			 * Pairs with mb in set_shrinker_bit():
> +			 *
> +			 * list_lru_add()     shrink_slab_memcg()
> +			 *   list_add_tail()    clear_bit()
> +			 *   <MB>               <MB>
> +			 *   set_bit()          do_shrink_slab()
> +			 */
> +			smp_mb__after_atomic();
> +			ret = do_shrink_slab(&sc, shrinker, priority);
> +			if (ret == SHRINK_EMPTY)
> +				ret = 0;
> +			else
> +				set_shrinker_bit(memcg, nid, i);
> +		}

This is mind-boggling. Are there any alternatives? For instance, can't
we clear the bit in list_lru_del, when we hold the list lock?

>  		freed += ret;
>  
>  		if (rwsem_is_contended(&shrinker_rwsem)) {
> 

  reply	other threads:[~2018-04-22 18:21 UTC|newest]

Thread overview: 37+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-04-17 18:52 [PATCH v2 00/12] Improve shrink_slab() scalability (old complexity was O(n^2), new is O(n)) Kirill Tkhai
2018-04-17 18:53 ` [PATCH v2 01/12] mm: Assign id to every memcg-aware shrinker Kirill Tkhai
2018-04-18 14:14   ` Tetsuo Handa
2018-04-18 14:27     ` Kirill Tkhai
2018-04-18 14:32       ` Tetsuo Handa
2018-04-18 15:02         ` Kirill Tkhai
2018-04-22 17:16   ` Vladimir Davydov
2018-04-17 18:53 ` [PATCH v2 02/12] memcg: Refactoring in mem_cgroup_alloc() Kirill Tkhai
2018-04-17 18:53 ` [PATCH v2 03/12] memcg: Refactoring in alloc_mem_cgroup_per_node_info() Kirill Tkhai
2018-04-17 18:53 ` [PATCH v2 04/12] mm: Assign memcg-aware shrinkers bitmap to memcg Kirill Tkhai
2018-04-18 12:55   ` kbuild test robot
2018-04-18 13:05     ` Kirill Tkhai
2018-04-22 17:59   ` Vladimir Davydov
2018-04-23 10:54     ` Kirill Tkhai
2018-04-24 11:28       ` Vladimir Davydov
2018-04-24 11:38         ` Kirill Tkhai
2018-04-24 12:15           ` Vladimir Davydov
2018-04-24 12:24             ` Kirill Tkhai
2018-04-28 15:08               ` Vladimir Davydov
2018-05-03 11:15                 ` Kirill Tkhai
2018-04-24 12:13         ` Kirill Tkhai
2018-04-23 11:02     ` Kirill Tkhai
2018-04-23 11:06     ` Kirill Tkhai
2018-04-24 11:08       ` Vladimir Davydov
2018-04-17 18:53 ` [PATCH v2 05/12] fs: Propagate shrinker::id to list_lru Kirill Tkhai
2018-04-22 18:03   ` Vladimir Davydov
2018-04-17 18:53 ` [PATCH v2 06/12] list_lru: Add memcg argument to list_lru_from_kmem() Kirill Tkhai
2018-04-17 18:54 ` [PATCH v2 07/12] list_lru: Pass dst_memcg argument to memcg_drain_list_lru_node() Kirill Tkhai
2018-04-17 18:54 ` [PATCH v2 08/12] list_lru: Pass lru " Kirill Tkhai
2018-04-17 18:54 ` [PATCH v2 09/12] mm: Set bit in memcg shrinker bitmap on first list_lru item apearance Kirill Tkhai
2018-04-17 18:54 ` [PATCH v2 10/12] mm: Iterate only over charged shrinkers during memcg shrink_slab() Kirill Tkhai
2018-04-22 18:19   ` Vladimir Davydov
2018-04-17 18:54 ` [PATCH v2 11/12] mm: Add SHRINK_EMPTY shrinker methods return value Kirill Tkhai
2018-04-17 18:54 ` [PATCH v2 12/12] mm: Clear shrinker bit if there are no objects related to memcg Kirill Tkhai
2018-04-22 18:21   ` Vladimir Davydov [this message]
2018-04-23 10:01     ` Kirill Tkhai
2018-04-24 10:56       ` Vladimir Davydov

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=20180422182132.c4tqkyy4ojgi7l7q@esperanza \
    --to=vdavydov.dev@gmail.com \
    --cc=akpm@linux-foundation.org \
    --cc=aryabinin@virtuozzo.com \
    --cc=chris@chris-wilson.co.uk \
    --cc=gregkh@linuxfoundation.org \
    --cc=guro@fb.com \
    --cc=hannes@cmpxchg.org \
    --cc=hillf.zj@alibaba-inc.com \
    --cc=jbacik@fb.com \
    --cc=ktkhai@virtuozzo.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=linux@roeck-us.net \
    --cc=lirongqing@baidu.com \
    --cc=longman@redhat.com \
    --cc=mgorman@techsingularity.net \
    --cc=mhocko@kernel.org \
    --cc=minchan@kernel.org \
    --cc=mka@chromium.org \
    --cc=penguin-kernel@I-love.SAKURA.ne.jp \
    --cc=pombredanne@nexb.com \
    --cc=sfr@canb.auug.org.au \
    --cc=shakeelb@google.com \
    --cc=stummala@codeaurora.org \
    --cc=tglx@linutronix.de \
    --cc=viro@zeniv.linux.org.uk \
    --cc=willy@infradead.org \
    --cc=ying.huang@intel.com \
    --subject='Re: [PATCH v2 12/12] mm: Clear shrinker bit if there are no objects related to memcg' \
    /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

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).