From mboxrd@z Thu Jan 1 00:00:00 1970 From: KAMEZAWA Hiroyuki Subject: Re: [PATCH 06/10] cgroup: convert memcg controller to the new cftype interface Date: Wed, 21 Mar 2012 13:42:39 +0900 Message-ID: <4F695C3F.3000406@jp.fujitsu.com> References: <1331940963-15756-1-git-send-email-tj@kernel.org> <1331940963-15756-7-git-send-email-tj@kernel.org> <4F66B55A.8020307@jp.fujitsu.com> <20120319161052.GC11069@google.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20120319161052.GC11069-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: containers-bounces-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org Errors-To: containers-bounces-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org To: Tejun Heo Cc: ctalbott-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org, rni-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org, fweisbec-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org, containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org, linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, Michal Hocko , Hugh Dickins , Johannes Weiner , cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org List-Id: containers.vger.kernel.org (2012/03/20 1:10), Tejun Heo wrote: > Hello, > > On Mon, Mar 19, 2012 at 01:26:02PM +0900, KAMEZAWA Hiroyuki wrote: >>> +CGROUP_SUBSYS_CFTYPES_COND(mem_cgroup_subsys, memsw_cgroup_files, >>> + really_do_swap_account); >> >> >> I'm sorry but why you use really_do_swap_account instead of do_swap_account ? > > Because do_swap_account is initialized too late. > CGROUP_SUBSYS_CFTYPES() are processed via fs_initcall(). AFAICS, > do_swap_account may not have been initialized by then. Also, if memcg > as whole is disabled, it doesn't matter whether those files are > registered or not, right? We probably need some comment there tho. > I think creating file based on really_do_swap_account is wrong. (I'm sorry for this complication.) How about creating this file unconditionally ? Returning -ENOTSUP in memcg's handler will work well...anyway it's configured, no reason to show them is bad. Thanks, -Kame