Return-Path: Received: from fieldses.org ([173.255.197.46]:46558 "EHLO fieldses.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751789AbdEQPqx (ORCPT ); Wed, 17 May 2017 11:46:53 -0400 Date: Wed, 17 May 2017 11:46:48 -0400 From: "J. Bruce Fields" To: Gu Zheng Cc: jlayton@poochiereds.net, linux-nfs@vger.kernel.org, houtao1@huawei.com, Zhaohongjiang , miaoxie@huawei.com, Kinglong Mee Subject: Re: nfs+ext4,can not umount the ext4 mountpoint Message-ID: <20170517154648.GB30168@fieldses.org> References: <591AF71F.1040703@huawei.com> <20170516174215.GA15232@fieldses.org> <591BF94E.9000805@huawei.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <591BF94E.9000805@huawei.com> Sender: linux-nfs-owner@vger.kernel.org List-ID: On Wed, May 17, 2017 at 03:18:38PM +0800, Gu Zheng wrote: > and now we do not set the crossmnt ,so that we saw is the underlying directory . > the nfs mount have no relation to ext4 mountpoint, it can be umounted maybe better. > > can we add some limits to let it do not enter the nfsd_cross_mnt() in nfsd_lookup_dentry()if the crossmnt is not set? If the problem is inability to unmount a filesystem that isn't actually even exported, then it *might* be addressed by Kinglong Mee's patches: http://lkml.kernel.org/r/55ED9899.9010401@gmail.com I'm not sure what those patches still need--they may just be waiting for my review. --b