Received-SPF: pass (google.com: domain of linux-kernel+bounces-87311-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:4601:e00::3 as permitted sender) client-ip=2604:1380:4601:e00::3;
Precedence: bulk
MIME-Version: 1.0
References: <20240229121650.33983-1-thorsten.blum@toblux.com>
In-Reply-To: <20240229121650.33983-1-thorsten.blum@toblux.com>
From: Ryusuke Konishi <konishi.ryusuke@gmail.com>
Date: Fri, 1 Mar 2024 03:45:55 +0900
Message-ID: <CAKFNMomtp7ZwB0gmxoemp_ums4rqOSbfF2BMS6kX+LwtKYtvCg@mail.gmail.com>
Subject: Re: [PATCH] nilfs2: Use div64_ul() instead of do_div()
To: Thorsten Blum <thorsten.blum@toblux.com>
Cc: linux-nilfs@vger.kernel.org, linux-kernel@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Thu, Feb 29, 2024 at 9:19=E2=80=AFPM Thorsten Blum wrote:
>
> Fixes Coccinelle/coccicheck warning reported by do_div.cocci.
>
> Compared to do_div(), div64_ul() does not implicitly cast the divisor and
> does not unnecessarily calculate the remainder.
>
> Signed-off-by: Thorsten Blum <thorsten.blum@toblux.com>
> ---
>  fs/nilfs2/cpfile.c | 2 +-
>  fs/nilfs2/dat.c    | 2 +-
>  fs/nilfs2/ioctl.c  | 4 ++--
>  fs/nilfs2/sufile.c | 2 +-
>  4 files changed, 5 insertions(+), 5 deletions(-)
>
> diff --git a/fs/nilfs2/cpfile.c b/fs/nilfs2/cpfile.c
> index 39136637f715..bafbdca1a17d 100644
> --- a/fs/nilfs2/cpfile.c
> +++ b/fs/nilfs2/cpfile.c
> @@ -28,7 +28,7 @@ nilfs_cpfile_get_blkoff(const struct inode *cpfile, __u=
64 cno)
>  {
>         __u64 tcno =3D cno + NILFS_MDT(cpfile)->mi_first_entry_offset - 1=
;
>
> -       do_div(tcno, nilfs_cpfile_checkpoints_per_block(cpfile));
> +       tcno =3D div64_ul(tcno, nilfs_cpfile_checkpoints_per_block(cpfile=
));
>         return (unsigned long)tcno;
>  }
>
> diff --git a/fs/nilfs2/dat.c b/fs/nilfs2/dat.c
> index 9cf6ba58f585..df5324b0c0cd 100644
> --- a/fs/nilfs2/dat.c
> +++ b/fs/nilfs2/dat.c
> @@ -460,7 +460,7 @@ ssize_t nilfs_dat_get_vinfo(struct inode *dat, void *=
buf, unsigned int visz,
>                 kaddr =3D kmap_atomic(entry_bh->b_page);
>                 /* last virtual block number in this block */
>                 first =3D vinfo->vi_vblocknr;
> -               do_div(first, entries_per_block);
> +               first =3D div64_ul(first, entries_per_block);
>                 first *=3D entries_per_block;
>                 last =3D first + entries_per_block - 1;
>                 for (j =3D i, n =3D 0;
> diff --git a/fs/nilfs2/ioctl.c b/fs/nilfs2/ioctl.c
> index cfb6aca5ec38..f1a01c191cf5 100644
> --- a/fs/nilfs2/ioctl.c
> +++ b/fs/nilfs2/ioctl.c
> @@ -1111,7 +1111,7 @@ static int nilfs_ioctl_set_alloc_range(struct inode=
 *inode, void __user *argp)
>         segbytes =3D nilfs->ns_blocks_per_segment * nilfs->ns_blocksize;
>
>         minseg =3D range[0] + segbytes - 1;
> -       do_div(minseg, segbytes);
> +       minseg =3D div64_ul(minseg, segbytes);
>
>         if (range[1] < 4096)
>                 goto out;
> @@ -1120,7 +1120,7 @@ static int nilfs_ioctl_set_alloc_range(struct inode=
 *inode, void __user *argp)
>         if (maxseg < segbytes)
>                 goto out;
>
> -       do_div(maxseg, segbytes);
> +       maxseg =3D div64_ul(maxseg, segbytes);
>         maxseg--;
>
>         ret =3D nilfs_sufile_set_alloc_range(nilfs->ns_sufile, minseg, ma=
xseg);
> diff --git a/fs/nilfs2/sufile.c b/fs/nilfs2/sufile.c
> index 0a8119456c21..c02b523d9c7e 100644
> --- a/fs/nilfs2/sufile.c
> +++ b/fs/nilfs2/sufile.c
> @@ -48,7 +48,7 @@ nilfs_sufile_get_blkoff(const struct inode *sufile, __u=
64 segnum)
>  {
>         __u64 t =3D segnum + NILFS_MDT(sufile)->mi_first_entry_offset;
>
> -       do_div(t, nilfs_sufile_segment_usages_per_block(sufile));
> +       t =3D div64_ul(t, nilfs_sufile_segment_usages_per_block(sufile));
>         return (unsigned long)t;
>  }
>
> --
> 2.44.0
>

All of the fixes in this patch seem to be correct, but this doesn't
cover nilfs_resize_fs(), nilfs_max_segment_count(), and
nilfs_sb2_bad_offset(), which also have do_div() that doesn't use the
return value.

Did do_div.cocci make a difference ?

Thanks,
Ryusuke Konishi