From mboxrd@z Thu Jan 1 00:00:00 1970 From: Alex Elder Subject: [PATCH 0/7] rbd: isolate mapping info, rearrange init Date: Fri, 07 Sep 2012 08:40:33 -0500 Message-ID: <5049F951.2030706@inktank.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Return-path: Received: from mail-ie0-f174.google.com ([209.85.223.174]:33245 "EHLO mail-ie0-f174.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754295Ab2IGNkf (ORCPT ); Fri, 7 Sep 2012 09:40:35 -0400 Received: by ieje11 with SMTP id e11so5004351iej.19 for ; Fri, 07 Sep 2012 06:40:34 -0700 (PDT) Sender: ceph-devel-owner@vger.kernel.org List-ID: To: ceph-devel@vger.kernel.org This third series begins a little more interesting code restructuring. The first five patches make more clear the distinction in a struct rbd_device between information about the underlying rbd image and information about the mapped snapshot represented by the structure. The last two patches (which might arguably belong in a later sub-series) move some initialization code out of one function and into its caller. It is available as branch "wip-rbd-review-3" on the ceph-client git repository, and is based on the branch "wip-rbd-review-2". https://github.com/ceph/ceph-client/tree/wip-rbd-review-3 -Alex [PATCH 1/7] f4069bf rbd: separate mapping info in rbd_dev This groups some related fields of an rbd device structure into a sub-structure. The fields pertain to the specific mapping an rbd_device represents--as opposed to the base rbd image. [PATCH 2/7] bee627f rbd: record mapped size This adds the size of the image or snapshot that's mapped to an rbd device's mapping information. [PATCH 3/7] 8b45382 rbd: return snap name from rbd_add_parse_args() [PATCH 4/7] cb93cc4 rbd: set mapping name with the rest These two patches make the name of the mapped snapshot be set along with the other mapping-related fields. [PATCH 5/7] 8a245ce rbd: simplify snap_by_name() interface This does some refactoring enabled by the previous few patches. [PATCH 6/7] e4008bd rbd: do some header initialization earlier This rearranges where certain initialization is done, beginning the process of re-ordering some of these steps in order to allow both format 1 and format 2 rbd images to be handled in the same way. [PATCH 7/7] 275ba12 rbd: simplify rbd_init_disk() a bit This does some refactoring enabled by the previous patch.