From: Peter Staubach Subject: Re: [PATCH 2/4] NFS: Support raw IPv6 address hostnames during NFS mount operation Date: Wed, 18 Jun 2008 18:59:03 -0400 Message-ID: <48599337.1070502@redhat.com> References: <20080618222951.16006.3679.stgit@ellison.1015granger.net> <20080618223203.16006.61765.stgit@ellison.1015granger.net> <48598DB7.4060404@redhat.com> <48598F59.2040505@oracle.com> <1213829515.25182.78.camel@localhost> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Cc: chuck.lever@oracle.com, linux-nfs@vger.kernel.org To: Trond Myklebust Return-path: Received: from mx1.redhat.com ([66.187.233.31]:32784 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754299AbYFRXAS (ORCPT ); Wed, 18 Jun 2008 19:00:18 -0400 In-Reply-To: <1213829515.25182.78.camel@localhost> Sender: linux-nfs-owner@vger.kernel.org List-ID: Trond Myklebust wrote: > On Wed, 2008-06-18 at 18:42 -0400, Chuck Lever wrote: > >> Peter Staubach wrote: >> >>> It seems unfortunate that the convention couldn't have been to >>> look for the first instance of ":/" and break the strings there. >>> >> I wonder if you can ever specify an export path that does not begin with >> a '/' -- is that possible when mounting an NFS server that exports, say, >> an old-fashioned HFS volume? >> > > It is quite legal to specify a path that doesn't start with '/': > NFSv2/v3 mount servers will, for instance, just implicitly assume it. > For that reason, there were user complaints the last time we started > requiring the '/' and returning an error when it wasn't supplied. And, hence, the [] syntax... :-) ps