Return-Path: linux-nfs-owner@vger.kernel.org Received: from fieldses.org ([174.143.236.118]:39058 "EHLO fieldses.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753029Ab1K1RGr (ORCPT ); Mon, 28 Nov 2011 12:06:47 -0500 Date: Mon, 28 Nov 2011 12:06:44 -0500 From: "J. Bruce Fields" To: Steve Dickson Cc: Tigran Mkrtchyan , linux-nfs Subject: Re: nfs-utils and libnfsidmap Message-ID: <20111128170644.GA3105@fieldses.org> References: <20111123201345.GA1044@fieldses.org> <4ED39F2D.7050508@RedHat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <4ED39F2D.7050508@RedHat.com> Sender: linux-nfs-owner@vger.kernel.org List-ID: On Mon, Nov 28, 2011 at 09:48:13AM -0500, Steve Dickson wrote: > On 11/23/2011 03:13 PM, J. Bruce Fields wrote: > > On Wed, Nov 23, 2011 at 08:54:06PM +0100, Tigran Mkrtchyan wrote: > >> Hi All, > >> > >> today was tracing a bug in nfs-utils/libnfsidmap and after a log > >> debugging located it and > >> finally realized that it's already fixed by Steve Dickson ( > >> http://git.linux-nfs.org/?p=steved/nfs-utils.git;a=commit;h=d22ef3f525d71b565fcc688557273a6cabeeb71a > >> ). Nevertheless during > >> this procedure it turned out that there is quite some code duplicated > >> between nfs-utils and > >> libnfsidmap. > >> > >> Questions: > >> a) why nfs-utils duplicates some part of libnfsidmap and still depends o it > >> b) is there readon for code duplication? Licensing or so > >> c) what ww need to do to get rid of duplication > >> > >> I expect the answer of 'c' will contain something like time and man power. > >> I am volunteering to to pend some time on it. > > > > I doubt there's any real reason for duplication. > Just curious as to what code we are talking about... Yes, maybe Tigran can demonstrate with a patch. > > Probably libnfsidmap should be part of nfs-utils, actually. > I guess we could roll the libnfsidmap git tree into the nfs-utils > tree... if that make senses... It probably would simply things.. OK. > > And maybe we don't need it at all--the original reason to split out > > libnfsidmap was to share the code with libacl, so the posix getfacl > > command could do v4->posix acl mapping, but those patches never made it > > upstream. > So you don't think this will every happen? I don't know. If it's easy to keep it as a separate library, then maybe we should just in case. But for now it seems unlikely. --b.