From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-qk0-f171.google.com (mail-qk0-f171.google.com [209.85.220.171]) by kanga.kvack.org (Postfix) with ESMTP id 1FAE96B026C for ; Fri, 1 Apr 2016 08:30:29 -0400 (EDT) Received: by mail-qk0-f171.google.com with SMTP id i4so37189535qkc.3 for ; Fri, 01 Apr 2016 05:30:29 -0700 (PDT) Received: from EUR01-VE1-obe.outbound.protection.outlook.com (mail-ve1eur01on0093.outbound.protection.outlook.com. [104.47.1.93]) by mx.google.com with ESMTPS id u82si11780707qhu.121.2016.04.01.05.30.26 for (version=TLS1_2 cipher=ECDHE-RSA-AES128-SHA bits=128/128); Fri, 01 Apr 2016 05:30:27 -0700 (PDT) From: Vladimir Davydov Subject: [PATCH] mm: slub: replace kick_all_cpus_sync with synchronize_sched in kmem_cache_shrink Date: Fri, 1 Apr 2016 15:30:17 +0300 Message-ID: <1459513817-11853-1-git-send-email-vdavydov@virtuozzo.com> MIME-Version: 1.0 Content-Type: text/plain Sender: owner-linux-mm@kvack.org List-ID: To: Andrew Morton Cc: Peter Zijlstra , Christoph Lameter , Joonsoo Kim , Pekka Enberg , David Rientjes , Johannes Weiner , Michal Hocko , linux-mm@kvack.org, linux-kernel@vger.kernel.org When we call __kmem_cache_shrink on memory cgroup removal, we need to synchronize kmem_cache->cpu_partial update with put_cpu_partial that might be running on other cpus. Currently, we achieve that by using kick_all_cpus_sync, which works as a system wide memory barrier. Though fast it is, this method has a flow - it issues a lot of IPIs, which might hurt high performance or real-time workloads. To fix this, let's replace kick_all_cpus_sync with synchronize_sched. Although the latter one may take much longer to finish, it shouldn't be a problem in this particular case, because memory cgroups are destroyed asynchronously from a workqueue so that no user visible effects should be introduced. OTOH, it will save us from excessive IPIs when someone removes a cgroup. Anyway, even if using synchronize_sched turns out to take too long, we can always introduce a kind of __kmem_cache_shrink batching so that this method would only be called once per one cgroup destruction (not per each per memcg kmem cache as it is now). Reported-and-suggested-by: Peter Zijlstra Signed-off-by: Vladimir Davydov --- mm/slub.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/mm/slub.c b/mm/slub.c index 279e773d80d3..03067f43dcf4 100644 --- a/mm/slub.c +++ b/mm/slub.c @@ -3697,7 +3697,7 @@ int __kmem_cache_shrink(struct kmem_cache *s, bool deactivate) * s->cpu_partial is checked locklessly (see put_cpu_partial), * so we have to make sure the change is visible. */ - kick_all_cpus_sync(); + synchronize_sched(); } flush_all(s); -- 2.1.4 -- 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: email@kvack.org