From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-fsdevel-owner@vger.kernel.org>
Received: from mail-lf0-f68.google.com ([209.85.215.68]:44856 "EHLO
        mail-lf0-f68.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
        with ESMTP id S1751284AbeFVOc1 (ORCPT
        <rfc822;linux-fsdevel@vger.kernel.org>);
        Fri, 22 Jun 2018 10:32:27 -0400
MIME-Version: 1.0
In-Reply-To: <20180620224553.wfz6abuvvz4ul42d@eaf>
References: <20180619160223.4108556-1-arnd@arndb.de> <20180619160223.4108556-3-arnd@arndb.de>
 <20180620224553.wfz6abuvvz4ul42d@eaf>
From: Arnd Bergmann <arnd@arndb.de>
Date: Fri, 22 Jun 2018 16:32:25 +0200
Message-ID: <CAK8P3a1io5zfDwZXV771cw3AzVu-gK4Vs2UAh9U2pQA=dtN_yg@mail.gmail.com>
Subject: Re: [PATCH 3/3] hfsplus: return inode birthtime for statx
To: =?UTF-8?Q?Ernesto_A=2E_Fern=C3=A1ndez?=
        <ernesto.mnd.fernandez@gmail.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>,
        Andrew Morton <akpm@linux-foundation.org>,
        y2038 Mailman List <y2038@lists.linaro.org>,
        Vyacheslav Dubeyko <slava@dubeyko.com>,
        Deepa Dinamani <deepa.kernel@gmail.com>,
        Jan Kara <jack@suse.cz>, Jeff Layton <jlayton@redhat.com>,
        Linux FS-devel Mailing List <linux-fsdevel@vger.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Sender: linux-fsdevel-owner@vger.kernel.org
List-ID: <linux-fsdevel.vger.kernel.org>

On Thu, Jun 21, 2018 at 12:45 AM, Ernesto A. Fern=C3=A1ndez
<ernesto.mnd.fernandez@gmail.com> wrote:

>> --- a/fs/hfsplus/inode.c
>> +++ b/fs/hfsplus/inode.c
>> @@ -276,6 +276,19 @@ static int hfsplus_setattr(struct dentry *dentry, s=
truct iattr *attr)
>>       return 0;
>>  }
>>
>> +int hfsplus_getattr(const struct path *path, struct kstat *stat,
>> +                  u32 request_mask, unsigned int query_flags)
>> +{
>> +     struct inode *inode =3D d_backing_inode(path->dentry);
>
> I think d_inode() is better. They work the same, but "normal filesystems
> should not use this", according to the d_backing_inode() documentation.
>

Right, definitely. I copied it from vfs_getattr_nosec() without thinking ab=
out
it much. I see how David Howells put that d_backing_inode() there, but
still don't understand it.

>> @@ -335,6 +348,7 @@ int hfsplus_file_fsync(struct file *file, loff_t sta=
rt, loff_t end,
>>
>>  static const struct inode_operations hfsplus_file_inode_operations =3D =
{
>>       .setattr        =3D hfsplus_setattr,
>> +     .getattr        =3D hfsplus_getattr,
>>       .listxattr      =3D hfsplus_listxattr,
>>  #ifdef CONFIG_HFSPLUS_FS_POSIX_ACL
>>       .get_acl        =3D hfsplus_get_posix_acl,
>> --
>> 2.9.0
>>
>
> What about symlinks and special files?

My mistake again, thanks for pointing that out. Doing the symlinks correctl=
y
here would actually add a bit more complexity as they use the generic
page_symlink_inode_operations at the moment.

I think I'd rather just retract this patch and let someone else handle it i=
f
they actually want this feature. I only added it because it seemed trivial
to do, but that was clearly not true. ;-)

       Arnd