Date: Tue, 30 May 2017 15:40:30 -0400
From: "J . Bruce Fields" <bfields@fieldses.org>
To: Stephen Smalley <sds@tycho.nsa.gov>
Cc: Scott Mayhew <smayhew@redhat.com>, linux-nfs@vger.kernel.org,
        selinux@tycho.nsa.gov, Trond Myklebust <trondmy@primarydata.com>
Subject: Re: [PATCH RFC 2/2] nfs: update labeling behavior on a superblock
 when submounting
Message-ID: <20170530194030.GB9371@fieldses.org>
References: <20170404232646.GB24146@parsley.fieldses.org>
 <20170525210754.24265-1-smayhew@redhat.com>
 <20170525210754.24265-3-smayhew@redhat.com>
 <1495808645.12091.10.camel@tycho.nsa.gov>
 <20170526152821.xgyidg7qebdwbm6f@tonberry.usersys.redhat.com>
 <1496155125.2164.8.camel@tycho.nsa.gov>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
In-Reply-To: <1496155125.2164.8.camel@tycho.nsa.gov>
Sender: linux-nfs-owner@vger.kernel.org

On Tue, May 30, 2017 at 10:38:45AM -0400, Stephen Smalley wrote:
> On Fri, 2017-05-26 at 11:28 -0400, Scott Mayhew wrote:
> > On Fri, 26 May 2017, Stephen Smalley wrote:
> > 
> > > On Thu, 2017-05-25 at 17:07 -0400, Scott Mayhew wrote:
> > > > When the client traverses from filesystem exported without the
> > > > "security_label" option to one exported with the "security_label"
> > > > option, it needs to pass SECURITY_LSM_NATIVE_LABELS to
> > > > security_sb_set_mnt_opts() so that the new superblock has
> > > > SBLABEL_MNT
> > > > set in its security mount options.  Otherwise, attempts to set
> > > > security
> > > > labels via setxattr over NFSv4.2 will fail.
> > > > 
> > > > Signed-off-by: Scott Mayhew <smayhew@redhat.com>
> > > > ---
> > > >  fs/nfs/super.c | 23 ++++++++++++++++++++++-
> > > >  1 file changed, 22 insertions(+), 1 deletion(-)
> > > > 
> > > > diff --git a/fs/nfs/super.c b/fs/nfs/super.c
> > > > index 2f3822a..d7a3b89 100644
> > > > --- a/fs/nfs/super.c
> > > > +++ b/fs/nfs/super.c
> > > > @@ -2544,10 +2544,31 @@ EXPORT_SYMBOL_GPL(nfs_set_sb_security);
> > > >  int nfs_clone_sb_security(struct super_block *s, struct dentry
> > > > *mntroot,
> > > >  			  struct nfs_mount_info *mount_info)
> > > >  {
> > > > +	int error;
> > > > +	unsigned long kflags = 0, kflags_out = 0;
> > > > +	struct security_mnt_opts opts;
> > > > +
> > > >  	/* clone any lsm security options from the parent to the
> > > > new
> > > > sb */
> > > >  	if (d_inode(mntroot)->i_op != NFS_SB(s)->nfs_client-
> > > > > rpc_ops->dir_inode_ops)
> > > > 
> > > >  		return -ESTALE;
> > > > -	return security_sb_clone_mnt_opts(mount_info->cloned-
> > > > >sb,
> > > > s);
> > > > +	error = security_sb_clone_mnt_opts(mount_info->cloned-
> > > > >sb,
> > > > s);
> > > > +	if (error)
> > > > +		goto err;
> > > > +
> > > > +	if (NFS_SB(s)->caps & NFS_CAP_SECURITY_LABEL &&
> > > > +		!(NFS_SB(mount_info->cloned->sb)->caps &
> > > > NFS_CAP_SECURITY_LABEL)) {
> > > > +		memset(&opts, 0, sizeof(opts));
> > > > +		kflags |= SECURITY_LSM_NATIVE_LABELS;
> > > > +
> > > > +		error = security_sb_set_mnt_opts(s, &opts,
> > > > kflags,
> > > > &kflags_out);
> > > > +		if (error)
> > > > +			goto err;
> > > > +
> > > > +		if (!(kflags_out & SECURITY_LSM_NATIVE_LABELS))
> > > > +			NFS_SB(s)->caps &=
> > > > ~NFS_CAP_SECURITY_LABEL;
> > > > +	}
> > > > +err:
> > > > +	return error;
> > > >  }
> > > >  EXPORT_SYMBOL_GPL(nfs_clone_sb_security);
> > > 
> > > Could this clobber a context set via context= mount option?
> > 
> > Argh, yes I suppose it could.  In my first attempt to fix this, I
> > added
> > a security_sb_get_mnt_opts() hook to get the original mount options
> > and
> > then passed that along with the SECURITY_LSM_NATIVE_LABELS flag to
> > security_sb_set_mnt_opts().  When I saw that
> > security_sb_set_mnt_opts()
> > wouldn't allow me to change a superblock that had already been
> > initialized, I got rid of the hook and added the check in patch 1...
> > maybe a combination of the two is needed?
> > 
> > Testing it again now, I'm not sure the context= mount option is
> > working
> > correctly with the latest kernel.
> 
> Looks like you are correct,
> https://github.com/SELinuxProject/selinux-kernel/issues/35

Ugh.  So, to make sure I understand: the desired behavior is that in the
case the client mounts with a context= option, behavior is exactly as if
the client or server didn't support the new security labeling protocol.
That would make sense to me.

--b.