Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <a6ca6c14a6b3727b7416198824b37bd7bd386180.camel@kernel.org>
Subject: Re: [PATCH v2 8/8] cifs: update the ctime on a partial page write
From:   Jeff Layton <jlayton@kernel.org>
To:     Tom Talpey <tom@talpey.com>,
        Christian Brauner <brauner@kernel.org>,
        Al Viro <viro@zeniv.linux.org.uk>,
        Brad Warrum <bwarrum@linux.ibm.com>,
        Ritu Agarwal <rituagar@linux.ibm.com>,
        Arnd Bergmann <arnd@arndb.de>,
        Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        Ian Kent <raven@themaw.net>,
        "Tigran A. Aivazian" <aivazian.tigran@gmail.com>,
        Jeremy Kerr <jk@ozlabs.org>, Ard Biesheuvel <ardb@kernel.org>,
        Namjae Jeon <linkinjeon@kernel.org>,
        Sungjong Seo <sj1557.seo@samsung.com>,
        Steve French <sfrench@samba.org>,
        Paulo Alcantara <pc@manguebit.com>,
        Ronnie Sahlberg <lsahlber@redhat.com>,
        Shyam Prasad N <sprasad@microsoft.com>,
        John Johansen <john.johansen@canonical.com>,
        Paul Moore <paul@paul-moore.com>,
        James Morris <jmorris@namei.org>,
        "Serge E. Hallyn" <serge@hallyn.com>,
        Ruihan Li <lrh2000@pku.edu.cn>,
        Sebastian Reichel <sebastian.reichel@collabora.com>,
        Alan Stern <stern@rowland.harvard.edu>,
        Suren Baghdasaryan <surenb@google.com>,
        Wolfram Sang <wsa+renesas@sang-engineering.com>,
        linux-kernel@vger.kernel.org, linux-usb@vger.kernel.org,
        autofs@vger.kernel.org, linux-efi@vger.kernel.org,
        linux-fsdevel@vger.kernel.org, linux-cifs@vger.kernel.org,
        samba-technical@lists.samba.org, apparmor@lists.ubuntu.com,
        linux-security-module@vger.kernel.org
Date:   Mon, 12 Jun 2023 09:51:50 -0400
In-Reply-To: <a34b598a-374b-5dbf-dd36-4b62e52fe36c@talpey.com>
References: <20230612104524.17058-1-jlayton@kernel.org>
         <20230612104524.17058-9-jlayton@kernel.org>
         <a34b598a-374b-5dbf-dd36-4b62e52fe36c@talpey.com>
Content-Type: text/plain; charset="ISO-8859-15"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.48.3 (3.48.3-1.fc38) 
MIME-Version: 1.0
Precedence: bulk

On Mon, 2023-06-12 at 09:41 -0400, Tom Talpey wrote:
> On 6/12/2023 6:45 AM, Jeff Layton wrote:
> > POSIX says:
> >=20
> >      "Upon successful completion, where nbyte is greater than 0, write(=
)
> >       shall mark for update the last data modification and last file st=
atus
> >       change timestamps of the file..."
> >=20
> > Add the missing ctime update.
> >=20
> > Signed-off-by: Jeff Layton <jlayton@kernel.org>
> > ---
> >   fs/smb/client/file.c | 2 +-
> >   1 file changed, 1 insertion(+), 1 deletion(-)
> >=20
> > diff --git a/fs/smb/client/file.c b/fs/smb/client/file.c
> > index df88b8c04d03..a00038a326cf 100644
> > --- a/fs/smb/client/file.c
> > +++ b/fs/smb/client/file.c
> > @@ -2596,7 +2596,7 @@ static int cifs_partialpagewrite(struct page *pag=
e, unsigned from, unsigned to)
> >   					   write_data, to - from, &offset);
> >   		cifsFileInfo_put(open_file);
> >   		/* Does mm or vfs already set times? */
> > -		inode->i_atime =3D inode->i_mtime =3D current_time(inode);
> > +		inode->i_atime =3D inode->i_mtime =3D inode->i_ctime =3D current_tim=
e(inode);
>=20
> Question. It appears that roughly half the filesystems in this series
> don't touch the i_atime in this case. And the other half do. Which is
> correct? Did they incorrectly set i_atime instead of i_ctime?
>=20

I noticed that too, and with this set, I decided to not make any atime
changes since I wasn't sure.

I think the answer to your question is "it depends". atime is supposed
to be updated on reads, not writes, but sometimes a write requires a RMW
cycle of some flavor so one can imagine that in some cases we'd need to
update all three.

In this case, I'm not sure that updating any of these times is the right
thing to do. This is called from ->launder_folio, so the syscall that
issued the write is long gone and we're in writeback here.

With NFS, we generally leave timestamp updates to the server. Should any
of these timestamps be updated by the (SMB1) client here?
--=20
Jeff Layton <jlayton@kernel.org>