From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753873Ab0LCSSA (ORCPT ); Fri, 3 Dec 2010 13:18:00 -0500 Received: from rcsinet10.oracle.com ([148.87.113.121]:51920 "EHLO rcsinet10.oracle.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752455Ab0LCSR6 convert rfc822-to-8bit (ORCPT ); Fri, 3 Dec 2010 13:17:58 -0500 Subject: Re: [PATCH] NFS: suppressing showing of default mount port value in /proc fixed Mime-Version: 1.0 (Apple Message framework v1082) Content-Type: text/plain; charset=utf-8 From: Chuck Lever In-Reply-To: <4CF9307D.9040504@parallels.com> Date: Fri, 3 Dec 2010 13:17:19 -0500 Cc: "Trond.Myklebust@netapp.com" , "linux-nfs@vger.kernel.org" , "linux-kernel@vger.kernel.org" Content-Transfer-Encoding: 8BIT Message-Id: References: <20101203170138.22955.86546.stgit@localhost6.localdomain6> <7AF6FF7E-F57C-493B-8880-25509CC54EDE@oracle.com> <4CF9307D.9040504@parallels.com> To: Kinsbursky Stanislav X-Mailer: Apple Mail (2.1082) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Dec 3, 2010, at 1:01 PM, Kinsbursky Stanislav wrote: > 03.12.2010 20:35, Chuck Lever пишет: >> >> On Dec 3, 2010, at 12:11 PM, Stanislav Kinsbursky wrote: >> >>> Default value for mount server port is set to NFS_UNSPEC_PORT (-1) and will not >>> be changed during parsing mount options for mound data version 6. This default >>> value will be showed for mountport in /proc/mounts always since current default >>> check is for zero value. This small mistake leads to big problem, because >>> during umount.nfs execution from old user-space utils (at least nfs-utils >>> 1.0.9) this value will be used as the server port to connect to. This request >>> will be rejected (since port is 65535) and thus nfs mount point can't be >>> unmounted. >> >> Note: this is only possible if /etc/mtab is a link to /proc/mounts. Not all systems have this configuration. >> > > I found it on CentOs 5.5 with RHEL6 kernel. > Actually, the current patch has a specific flaw: if mount port is not passed by nfs utils using mount options version 6, then it will be set to 0 and showed in /proc/mounts as "mountport=0". Right: zero and negative one are both special internal port values that should not be displayed in /proc/mounts. Why not check if mountport > 0 instead? > Another fix solution is to call "nfs_set_port(sap, &args->nfs_server.port, 0);" when parsing mount options version 6 (like in done in default casein nfs_validate_mount_data) instead of changing nfs_show_mountd_options(). I'm not sure I follow here. Why would setting nfsport be the right thing to do? The default case is for text-based mounts only. > Kernel version affected: 2.6.37-rc4 >>> >>> Signed-off-by: Stanislav Kinsbursky >>> >>> --- >>> fs/nfs/super.c | 3 ++- >>> 1 files changed, 2 insertions(+), 1 deletions(-) >>> >>> diff --git a/fs/nfs/super.c b/fs/nfs/super.c >>> index 6d6e21d..fd4cac8 100644 >>> --- a/fs/nfs/super.c >>> +++ b/fs/nfs/super.c >>> @@ -681,7 +681,8 @@ static void nfs_show_mountd_options(struct seq_file *m, struct nfs_server *nfss, >>> >>> if (nfss->mountd_version || showdefaults) >>> seq_printf(m, ",mountvers=%u", nfss->mountd_version); >>> - if (nfss->mountd_port || showdefaults) >>> + if (nfss->mountd_port != (unsigned short)NFS_UNSPEC_PORT || >>> + showdefaults) >>> seq_printf(m, ",mountport=%u", nfss->mountd_port); >>> >>> nfs_show_mountd_netid(m, nfss, showdefaults); >>> >>> -- >>> To unsubscribe from this list: send the line "unsubscribe linux-nfs" in >>> the body of a message to majordomo@vger.kernel.org >>> More majordomo info at http://vger.kernel.org/majordomo-info.html >> > > > -- > Best regards, > Stanislav Kinsbursky -- Chuck Lever chuck[dot]lever[at]oracle[dot]com