From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755194AbbDUPCi (ORCPT ); Tue, 21 Apr 2015 11:02:38 -0400 Received: from mail-qc0-f170.google.com ([209.85.216.170]:36160 "EHLO mail-qc0-f170.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750960AbbDUPCg (ORCPT ); Tue, 21 Apr 2015 11:02:36 -0400 Date: Tue, 21 Apr 2015 11:02:29 -0400 From: Tejun Heo To: Jan Kara Cc: axboe@kernel.dk, linux-kernel@vger.kernel.org, hch@infradead.org, hannes@cmpxchg.org, linux-fsdevel@vger.kernel.org, vgoyal@redhat.com, lizefan@huawei.com, cgroups@vger.kernel.org, linux-mm@kvack.org, mhocko@suse.cz, clm@fb.com, fengguang.wu@intel.com, david@fromorbit.com, gthelen@google.com, Miklos Szeredi , Trond Myklebust Subject: Re: [PATCH 12/49] writeback: move backing_dev_info->bdi_stat[] into bdi_writeback Message-ID: <20150421150229.GA9455@htj.duckdns.org> References: <1428350318-8215-1-git-send-email-tj@kernel.org> <1428350318-8215-13-git-send-email-tj@kernel.org> <20150420150231.GA17020@quack.suse.cz> <20150420175626.GB4206@htj.duckdns.org> <20150421085119.GA24278@quack.suse.cz> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20150421085119.GA24278@quack.suse.cz> User-Agent: Mutt/1.5.23 (2014-03-12) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Tue, Apr 21, 2015 at 10:51:19AM +0200, Jan Kara wrote: > I can easily understand what "initializing writeback structure" means but > "exiting writeback structure" doesn't really make sense to me. OTOH > "destroying writeback structure" does make sense to me. That's the only > reason. We have enough cases where "exit" is used that way starting with module_exit() and all the accompanying __exit annotations and there are quite a few others. I think it's enough to establish "exit" as the counterpart of "init" but I do agree that it felt a bit alien to me at the beginning too. In general, I've been sticking with create/destroy if the object itself is being created or destroyed and init/exit if the object itself stays put across init/exit which is the case here. This isn't quite universal but I think there exists enough of a pattern to make it worthwhile to stick to it. As such, I'd like to stick to the current names if it isn't a big deal. Thanks. -- tejun