Return-Path: Received: from out2.smtp.messagingengine.com ([66.111.4.26]:59901 "EHLO out2.smtp.messagingengine.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751140Ab1IVPy0 (ORCPT ); Thu, 22 Sep 2011 11:54:26 -0400 Subject: Re: [PATCH] NFS4: Revert commit to make the automount code ignore LOOKUP_FOLLOW From: Ian Kent To: Linus Torvalds Cc: David Howells , miklos@szeredi.hu, viro@zeniv.linux.org.uk, jlayton@redhat.com, gregkh@suse.de, linux-nfs@vger.kernel.org, leonardo.lists@gmail.com In-Reply-To: <6490.1316704925@redhat.com> References: <20110922124608.7739.27750.stgit@warthog.procyon.org.uk> <6490.1316704925@redhat.com> Content-Type: text/plain; charset="UTF-8" Date: Thu, 22 Sep 2011 23:54:19 +0800 Message-ID: <1316706859.3346.37.camel@perseus.themaw.net> Sender: linux-nfs-owner@vger.kernel.org List-ID: MIME-Version: 1.0 On Thu, 2011-09-22 at 16:22 +0100, David Howells wrote: > > I would also say that we do not want lstat(), l*xattr() and co. to cause > automounting - but maybe they should. Perhaps if you don't want to cause > automounting, you should explicitly pass AT_NO_AUTOMOUNT, and all path-taking > VFS calls should have variants that accept this flag. I still think we had it right the first time. After all we have always said automounts are directories that have symlink semantics, which to my way of thinking is what we now have (or had). Not only that, it provides a natural way of ensuring a mount occurs or does not occur by using the well known no follow calls. The "natural" part of this is automounts can be followed conceptually in the same way that symlinks can be followed. I