From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from cn.fujitsu.com ([59.151.112.132]:32851 "EHLO heian.cn.fujitsu.com" rhost-flags-OK-FAIL-OK-FAIL) by vger.kernel.org with ESMTP id S1751335AbcCYBoi (ORCPT ); Thu, 24 Mar 2016 21:44:38 -0400 Subject: Re: [PATCH v8 25/27] btrfs: dedupe: Add support for compression and dedpue To: Chris Mason , , Wang Xiaoguang References: <1458610552-9845-1-git-send-email-quwenruo@cn.fujitsu.com> <1458610552-9845-26-git-send-email-quwenruo@cn.fujitsu.com> <20160324203504.m2b3lybguuf5r2hd@floor.thefacebook.com> From: Qu Wenruo Message-ID: <56F497FF.3010705@cn.fujitsu.com> Date: Fri, 25 Mar 2016 09:44:31 +0800 MIME-Version: 1.0 In-Reply-To: <20160324203504.m2b3lybguuf5r2hd@floor.thefacebook.com> Content-Type: text/plain; charset="utf-8"; format=flowed Sender: linux-btrfs-owner@vger.kernel.org List-ID: Chris Mason wrote on 2016/03/24 16:35 -0400: > On Tue, Mar 22, 2016 at 09:35:50AM +0800, Qu Wenruo wrote: >> From: Wang Xiaoguang >> >> The basic idea is also calculate hash before compression, and add needed >> members for dedupe to record compressed file extent. >> >> Since dedupe support dedupe_bs larger than 128K, which is the up limit >> of compression file extent, in that case we will skip dedupe and prefer >> compression, as in that size dedupe rate is low and compression will be >> more obvious. >> >> Current implement is far from elegant. The most elegant one should split >> every data processing method into its own and independent function, and >> have a unified function to co-operate them. > > I'd leave this one out for now, it looks like we need to refine the > pipeline from dedup -> compression and this is just more to carry around > until the initial support is in. Can you just decline to dedup > compressed extents for now? Yes, completely no problem. Although this patch seems works well yet, but I also have planned to rework current run_delloc_range() to make it more flex and clear. So the main object of the patch is more about raising attention for such further re-work. And now it has achieved its goal. Thanks, Qu > > -chris > >