From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-qg0-f49.google.com (mail-qg0-f49.google.com [209.85.192.49]) by kanga.kvack.org (Postfix) with ESMTP id CC66E6B0038 for ; Tue, 5 May 2015 09:45:27 -0400 (EDT) Received: by qgej70 with SMTP id j70so81131027qge.2 for ; Tue, 05 May 2015 06:45:27 -0700 (PDT) Received: from mail-qc0-x229.google.com (mail-qc0-x229.google.com. [2607:f8b0:400d:c01::229]) by mx.google.com with ESMTPS id i74si15773938qge.99.2015.05.05.06.45.25 for (version=TLSv1.2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128); Tue, 05 May 2015 06:45:26 -0700 (PDT) Received: by qcyk17 with SMTP id k17so87966600qcy.1 for ; Tue, 05 May 2015 06:45:25 -0700 (PDT) Date: Tue, 5 May 2015 09:45:21 -0400 From: Tejun Heo Subject: Re: [PATCH 2/2] kernfs: do not account ino_ida allocations to memcg Message-ID: <20150505134521.GL1971@htj.duckdns.org> References: <0cf48f4219721952f182715a61910f626d7c4aca.1430819044.git.vdavydov@parallels.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <0cf48f4219721952f182715a61910f626d7c4aca.1430819044.git.vdavydov@parallels.com> Sender: owner-linux-mm@kvack.org List-ID: To: Vladimir Davydov Cc: Andrew Morton , Johannes Weiner , Michal Hocko , Christoph Lameter , Pekka Enberg , David Rientjes , Joonsoo Kim , Greg Thelen , linux-mm@kvack.org, cgroups@vger.kernel.org, linux-kernel@vger.kernel.org On Tue, May 05, 2015 at 12:45:43PM +0300, Vladimir Davydov wrote: > root->ino_ida is used for kernfs inode number allocations. Since IDA has > a layered structure, different IDs can reside on the same layer, which > is currently accounted to some memory cgroup. The problem is that each > kmem cache of a memory cgroup has its own directory on sysfs (under > /sys/fs/kernel//cgroup). If the inode number of such a > directory or any file in it gets allocated from a layer accounted to the > cgroup which the cache is created for, the cgroup will get pinned for > good, because one has to free all kmem allocations accounted to a cgroup > in order to release it and destroy all its kmem caches. That said we > must not account layers of ino_ida to any memory cgroup. > > Since per net init operations may create new sysfs entries directly > (e.g. lo device) or indirectly (nf_conntrack creates a new kmem cache > per each namespace, which, in turn, creates new sysfs entries), an easy > way to reproduce this issue is by creating network namespace(s) from > inside a kmem-active memory cgroup. > > Signed-off-by: Vladimir Davydov Man, that's nasty. For the kernfs part, Acked-by: Tejun Heo Can you please repost this patch w/ Greg KH cc'd? Thanks. -- tejun -- 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