From mboxrd@z Thu Jan 1 00:00:00 1970 From: Amir Goldstein Subject: Re: [PATCH v5 1/4] ovl: move include of ovl_entry.h into overlayfs.h Date: Tue, 31 Oct 2017 15:22:45 +0200 Message-ID: References: <1509395247-15180-1-git-send-email-amir73il@gmail.com> <1509395247-15180-2-git-send-email-amir73il@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Return-path: Received: from mail-yw0-f193.google.com ([209.85.161.193]:48145 "EHLO mail-yw0-f193.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752589AbdJaNWq (ORCPT ); Tue, 31 Oct 2017 09:22:46 -0400 Received: by mail-yw0-f193.google.com with SMTP id q1so14698891ywh.5 for ; Tue, 31 Oct 2017 06:22:46 -0700 (PDT) In-Reply-To: Sender: linux-unionfs-owner@vger.kernel.org List-Id: linux-unionfs@vger.kernel.org To: Miklos Szeredi Cc: Chandan Rajendra , zhangyi , "linux-unionfs@vger.kernel.org" On Tue, Oct 31, 2017 at 3:14 PM, Miklos Szeredi wrote: > Is this really necessary? > > "ovl_entry.h" is supposed to be more of a low level thing, and I'm not > sure having that proliferate to all parts of the code is a good idea. > > At least some justification would be good here. The "problem" is that include "ovl_entry.h" is already included by 5/8 overlayfs c files and by the end of the NFS export series its 7/9 and there are exported functions in overlayfs.h whose signature include pointers to structs defined in ovl_entry.h, so I started forward declaring those structs in overlayfs.h. After you look at the next patch "anonymous devs for lowerdirs" let me know if you prefer a different distribution between the 2 include files, or a 3rd include file. Thanks, Amir.