Return-Path: <linux-nfs-owner@vger.kernel.org>
Received: from mail-px0-f173.google.com ([209.85.212.173]:52636 "EHLO
	mail-px0-f173.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
	with ESMTP id S1756750Ab1EKVfE convert rfc822-to-8bit (ORCPT
	<rfc822;linux-nfs@vger.kernel.org>); Wed, 11 May 2011 17:35:04 -0400
In-Reply-To: <1305148125.21868.2.camel@lade.trondhjem.org>
References: <1305147805-5756-1-git-send-email-shawn.p.huang@gmail.com>
	<1305148125.21868.2.camel@lade.trondhjem.org>
Date: Wed, 11 May 2011 17:35:03 -0400
Message-ID: <BANLkTimWE8FYGh9b7n9QF452dR0tRff50A@mail.gmail.com>
Subject: Re: [PATCH] nfs: check a crash in nfs_lookup_revalidate
From: Peng Huang <shawn.p.huang@gmail.com>
To: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: linux-nfs@vger.kernel.org, linux-kernel@vger.kernel.org
Content-Type: text/plain; charset=UTF-8
Sender: linux-nfs-owner@vger.kernel.org
List-ID: <linux-nfs.vger.kernel.org>
MIME-Version: 1.0

On Wed, May 11, 2011 at 5:08 PM, Trond Myklebust
<Trond.Myklebust@netapp.com> wrote:
> On Wed, 2011-05-11 at 17:03 -0400, Peng Huang wrote:
>> lookup_one_len() may call nfs_loopup_revalidate() with nd == NULL
>> indirectly, that causes the kernel crash.
>>
>> RIP: 0010:[<ffffffffa0ba3b41>]  [<ffffffffa0ba3b41>]
>> nfs_lookup_revalidate+0x21/0x4a0 [nfs]
>> RSP: 0018:ffff88018f00fae8  EFLAGS: 00010286
>>
>> Call Trace:
>>  [<ffffffff81164a17>] do_revalidate+0x17/0x60
>>  [<ffffffff81164e9b>] __lookup_hash+0xcb/0x140
>>  [<ffffffff811653c4>] lookup_one_len+0x94/0xe0
>>  [<ffffffff81241ef1>] ecryptfs_lookup+0x91/0x1d0
>>  [<ffffffff81164d85>] d_alloc_and_lookup+0x45/0x90
>>  [<ffffffff8116f7b5>] ? d_lookup+0x35/0x60
>>  [<ffffffff811669b2>] do_lookup+0x192/0x2d0
>>  [<ffffffff811763be>] ? vfsmount_lock_local_unlock+0x1e/0x30
>>  [<ffffffff8126d09c>] ? security_inode_permission+0x1c/0x30
>>  [<ffffffff81167d67>] link_path_walk+0x597/0xae0
>>  [<ffffffff8117638e>] ? vfsmount_lock_local_lock+0x1e/0x30
>>  [<ffffffff81165905>] ? path_init_rcu+0xa5/0x210
>>  [<ffffffff8116858b>] do_path_lookup+0x5b/0x140
>>  [<ffffffff811692f7>] user_path_at+0x57/0xa0
>>  [<ffffffff8159fd08>] ? do_page_fault+0x1e8/0x4e0
>>  [<ffffffff8115eb86>] vfs_fstatat+0x46/0x80
>>  [<ffffffff8116b990>] ? filldir+0x0/0xe0
>>  [<ffffffff8115ec2e>] vfs_lstat+0x1e/0x20
>>  [<ffffffff8115ec54>] sys_newlstat+0x24/0x50
>>  [<ffffffff8159c995>] ? page_fault+0x25/0x30
>>  [<ffffffff8100bfc2>] system_call_fastpath+0x16/0x1b
>>
>> Signed-off-by: Peng Huang <shawn.p.huang@gmail.com>
>> ---
>>  fs/nfs/dir.c |    2 +-
>>  1 files changed, 1 insertions(+), 1 deletions(-)
>>
>> diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
>> index 2c3eb33..9452aa5 100644
>> --- a/fs/nfs/dir.c
>> +++ b/fs/nfs/dir.c
>> @@ -1028,7 +1028,7 @@ static int nfs_lookup_revalidate(struct dentry *dentry, struct nameidata *nd)
>>       struct nfs_fattr *fattr = NULL;
>>       int error;
>>
>> -     if (nd->flags & LOOKUP_RCU)
>> +     if (nd != NULL && nd->flags & LOOKUP_RCU)
>>               return -ECHILD;
>>
>>       parent = dget_parent(dentry);
>
> That's exactly what Tyler Hicks proposed last week and which was NACKed.
> We simply won't support layered filesystems that don't do intents.
>
> IOW: Feel free to change the above to.
>
> if (nd == NULL)
>     return -EIO;
>

I tested returning -EIO when nd is NULL. Kernel does not crash, but
ecryptfs can not work on nfs anymore.

Peng Huang

>
>
>
> --
> Trond Myklebust
> Linux NFS client maintainer
>
> NetApp
> Trond.Myklebust@netapp.com
> www.netapp.com
>
>