From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tejun Heo Subject: Re: [GIT PULL] cgroup changes for v3.15-rc1 Date: Fri, 4 Apr 2014 08:03:59 -0400 Message-ID: <20140404120359.GA5461@mtj.dyndns.org> References: <20140403164911.GE24119@htj.dyndns.org> <20140403194335.GC2472@mtj.dyndns.org> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=sender:date:from:to:cc:subject:message-id:references:mime-version :content-type:content-disposition:in-reply-to:user-agent; bh=xUma3S95denRAQEH/0BHt2uhcS1bxILB5oW9e2jDsbA=; b=HccPnE89DH3s/F73B96Py/vzQx8jKQY/ezuVmUrw6/ZsqJC785NjXi5YG+9c+um1NH WCcjGu8TS+Sljgy0j47xxjV7g4ysi/VIxh5xjO4C15V/aTnoUBb857r+xewIyDoNC3AD eDe3A+HFxeqeMI1xqb/RV5kJlAiJc4uCTemMrQ25t9aSDEx+n7ig0iyAPei02E2cckmH uiN0SIMNpa6F08VUyir1SeX+CRD5P3evnQEnvoQ8ccYnW+2A/iFozU5ECBnciWC7dFLt g6WDjMFhk9hVzLv2hcKyG3CrcJsFfOq3Q7by2xIKAyk5AmP9zxuFga+mUBS7WBrl9aJr B/GQ== Content-Disposition: inline In-Reply-To: List-Id: 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: Linus Torvalds Cc: Linux Containers , Linux Kernel Mailing List , "Eric W. Biederman" , cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, David Miller On Thu, Apr 03, 2014 at 01:02:45PM -0700, Linus Torvalds wrote: > > The cgroup_root should be destroyed but it isn't, I think. We'd need > > to bump cgroup_root's refcnt only when a new sb is created. It's > > kinda ugly. Hmmm... > > Ok, so I guess we can use that "new_sb_created" thing, and I'll redo > my merge resolution to reflect that. I do find this incredibly ugly. I apparently missed the issue and designed the interface without considering this ugliness. Maybe kernfs could be made to wrap rather than providing mount/kill_sb() functions and hide details about sb or we can simply add fstype->umount() so that there's symmetry; however, the problem is kernfs-specific and other kernfs users would have single static backing store and won't need to care about this, so, for now, I think what it's a ugly but acceptable compromise. Thanks. -- tejun