From: Theodore Ts'o Subject: Re: [PATCH v3 2/6] ext4: track extent status tree shrinker delay statictics Date: Thu, 4 Sep 2014 11:49:25 -0400 Message-ID: <20140904154925.GF4047@thunk.org> References: <1407382553-24256-1-git-send-email-wenqing.lz@taobao.com> <1407382553-24256-3-git-send-email-wenqing.lz@taobao.com> <20140827132607.GA22211@quack.suse.cz> <20140904121048.GA13193@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii To: Jan Kara , linux-ext4@vger.kernel.org, Andreas Dilger , Zheng Liu Return-path: Received: from imap.thunk.org ([74.207.234.97]:37951 "EHLO imap.thunk.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752913AbaIDPtb (ORCPT ); Thu, 4 Sep 2014 11:49:31 -0400 Content-Disposition: inline In-Reply-To: <20140904121048.GA13193@gmail.com> Sender: linux-ext4-owner@vger.kernel.org List-ID: On Thu, Sep 04, 2014 at 08:10:48PM +0800, Zheng Liu wrote: > > When reading through this again, I realized that we probably don't want > > this file in /proc/fs/ext4 but rather in /sys/kernel/debug/ext4 because > > it's really a debugging interface and we don't want any tools to start > > using it. > > Fair enough. I will fix it in next version. I was actually OK with leaving it in /proc, since we have a number of other debugging files in /proc/fs/ext4/ already, and I'm not sure spreading ext4's foot print so that we have some files in /proc, some in /sys, and some in debugfs is worth decreasing the possibility that some tool would start depending on the contents of that file --- I can't really imagine why non-fs developers would ever care about such statistics. I probably *would* be in favor of creating a new config option which disables /proc/fs/ext4//mballoc_groups and /proc/fs/ext4//es_shrinker_info, to help with the people who care shrinking the total footprint of the Linux kernel for embedded application (kernel tinification), though. - Ted