Received-SPF: pass (google.com: domain of linux-ext4-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <5290be64ba87d01938c578f49443ce41f9be5e77.camel@kernel.org>
Subject: Re: [PATCH v2 42/92] ext4: convert to ctime accessor functions
From:   Jeff Layton <jlayton@kernel.org>
To:     Jan Kara <jack@suse.cz>
Cc:     Christian Brauner <brauner@kernel.org>,
        Theodore Ts'o <tytso@mit.edu>,
        Andreas Dilger <adilger.kernel@dilger.ca>,
        Al Viro <viro@zeniv.linux.org.uk>,
        linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org,
        linux-ext4@vger.kernel.org
Date:   Thu, 06 Jul 2023 09:10:17 -0400
In-Reply-To: <20230706123643.3pumra5f4fthz3qq@quack3>
References: <20230705185755.579053-1-jlayton@kernel.org>
         <20230705190309.579783-1-jlayton@kernel.org>
         <20230705190309.579783-40-jlayton@kernel.org>
         <20230706123643.3pumra5f4fthz3qq@quack3>
Content-Type: text/plain; charset="ISO-8859-15"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.48.4 (3.48.4-1.fc38) 
MIME-Version: 1.0
Precedence: bulk

On Thu, 2023-07-06 at 14:36 +0200, Jan Kara wrote:
> On Wed 05-07-23 15:01:07, Jeff Layton wrote:
> > In later patches, we're going to change how the inode's ctime field is
> > used. Switch to using accessor functions instead of raw accesses of
> > inode->i_ctime.
> >=20
> > Signed-off-by: Jeff Layton <jlayton@kernel.org>
>=20
> Some comment below:
>=20
> > diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
> > index 0a2d55faa095..d502b930431b 100644
> > --- a/fs/ext4/ext4.h
> > +++ b/fs/ext4/ext4.h
> > @@ -3823,6 +3823,27 @@ static inline int ext4_buffer_uptodate(struct bu=
ffer_head *bh)
> >  	return buffer_uptodate(bh);
> >  }
> > =20
> > +static inline void ext4_inode_set_ctime(struct inode *inode, struct ex=
t4_inode *raw_inode)
> > +{
> > +	struct timespec64 ctime =3D inode_get_ctime(inode);
> > +
> > +	if (EXT4_FITS_IN_INODE(raw_inode, EXT4_I(inode), i_ctime_extra)) {
> > +		raw_inode->i_ctime =3D cpu_to_le32(ctime.tv_sec);
> > +		raw_inode->i_ctime_extra =3D ext4_encode_extra_time(&ctime);
> > +	} else {
> > +		raw_inode->i_ctime =3D cpu_to_le32(clamp_t(int32_t, ctime.tv_sec, S3=
2_MIN, S32_MAX));
> > +	}
> > +}
> > +
> > +static inline void ext4_inode_get_ctime(struct inode *inode, const str=
uct ext4_inode *raw_inode)
> > +{
> > +	struct timespec64 ctime =3D { .tv_sec =3D (signed)le32_to_cpu(raw_ino=
de->i_ctime) };
> > +
> > +	if (EXT4_FITS_IN_INODE(raw_inode, EXT4_I(inode), i_ctime_extra))
> > +		ext4_decode_extra_time(&ctime, raw_inode->i_ctime_extra);
> > +	inode_set_ctime(inode, ctime.tv_sec, ctime.tv_nsec);
> > +}
> > +
>=20
> This duplication is kind of unpleasant. I was looking into it for a while
> and I think we can rather do some initial cleanup (attached patch 1) and
> then your conversion patch would not need to duplicate the conversion cod=
e
> (see attached patch 2).
>=20
> =09
>=20

Thanks Jan. That looks fine at first glance. I'll plan to drop my ext4
patch and replace it with these.

Cheers,
--=20
Jeff Layton <jlayton@kernel.org>