From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Sat, 11 May 2002 05:30:08 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Sat, 11 May 2002 05:30:07 -0400 Received: from imladris.infradead.org ([194.205.184.45]:48140 "EHLO phoenix.infradead.org") by vger.kernel.org with ESMTP id ; Sat, 11 May 2002 05:30:06 -0400 Date: Sat, 11 May 2002 10:23:48 +0100 From: Christoph Hellwig To: Alexander Viro Cc: Jan Harkes , linux-kernel@vger.kernel.org, kaos@ocs.com.au Subject: Re: [PATCH] iget-locked [2/6] Message-ID: <20020511102348.B24305@infradead.org> Mail-Followup-To: Christoph Hellwig , Alexander Viro , Jan Harkes , linux-kernel@vger.kernel.org, kaos@ocs.com.au In-Reply-To: <20020511030437.GA29392@ravel.coda.cs.cmu.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5.1i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Fri, May 10, 2002 at 11:44:23PM -0400, Alexander Viro wrote: > As for the binary compatibility... as long as we are source-compatible > (i.e. keep ->read_inode2 and provide a compatible iget4()) - compatibility > is not a problem. Anyone who ships binary-only modules is playing in the > traffic and if they become a roadkill - it's Not Our Problem(tm). Think > of it as evolution in action... We need to keep iget4, but I don't think keeping ->read_inode2 source compatiblitly is needed, it always was marked as a reiserfs cludge and no one was supposed to use it.