Received-SPF: pass (google.com: domain of linux-kernel+bounces-51383-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45d1:ec00::1 as permitted sender) client-ip=2604:1380:45d1:ec00::1;
Precedence: bulk
MIME-Version: 1.0
References: <20240203152436.1352-1-qwjhust@gmail.com> <8eaf59a4-1aaa-460e-a3cc-b798ed5e0f63@kernel.org>
 <CAGFpFsTz_9Zaj0PuptjBxOJwxF68geAUfd1qtx9--Tczh+jZww@mail.gmail.com> <e42ae14e-88b8-41de-9693-1513bf335fc2@kernel.org>
In-Reply-To: <e42ae14e-88b8-41de-9693-1513bf335fc2@kernel.org>
From: Wenjie Qi <qwjhust@gmail.com>
Date: Sun, 4 Feb 2024 10:36:15 +0800
Message-ID: <CAGFpFsTwH8S5QsDCd5hqS+WNNr4ZZPJ2FuPFLRWzkVCqfjxiQQ@mail.gmail.com>
Subject: Re: [PATCH v4] f2fs: fix zoned block device information initialization
To: Chao Yu <chao@kernel.org>
Cc: jaegeuk@kernel.org, yangyongpeng1@oppo.com, 
	linux-f2fs-devel@lists.sourceforge.net, linux-kernel@vger.kernel.org, 
	hustqwj@hust.edu.cn
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

I agree with you, it looks a bit more intuitive that way.
I'll submit a new version of the patch as soon as I can.
Thanks,

On Sun, Feb 4, 2024 at 10:24=E2=80=AFAM Chao Yu <chao@kernel.org> wrote:
>
> On 2024/2/4 10:18, Wenjie Qi wrote:
> > Hi Chao,
> >
> > It seems to me that when mounting multiple zoned devices,
> > if their max_open_zones are all 0, then sbi->max_open_zones is 0.
> > This suggests that all of the mounted devices can open an unlimited
> > number of zones,
> > and that we don't need to compare sbi->max_open_zones with
> > F2FS_OPTION( sbi).active_logs.
>
> Yes, but I'm curious about how this case (sbi->max_open_zones is zero)
> works w/ following patch, do we need to initialized sbi->max_open_zones
> w/ UINT_MAX to indicate the unlimited open zone status of device if
> all zoned devices' max_open_zones is zero?
>
> >
> > Thanks,
> >
> > Chao Yu <chao@kernel.org> =E4=BA=8E2024=E5=B9=B42=E6=9C=884=E6=97=A5=E5=
=91=A8=E6=97=A5 09:47=E5=86=99=E9=81=93=EF=BC=9A
> >>
> >> On 2024/2/3 23:24, Wenjie Qi wrote:
> >>> If the max open zones of zoned devices are less than
> >>> the active logs of F2FS, the device may error due to
> >>> insufficient zone resources when multiple active logs are
> >>> being written at the same time. If this value is 0,
> >>> there is no limit.
> >>>
> >>> Signed-off-by: Wenjie Qi <qwjhust@gmail.com>
> >>> ---
> >>>    fs/f2fs/f2fs.h  |  1 +
> >>>    fs/f2fs/super.c | 21 +++++++++++++++++++++
> >>>    2 files changed, 22 insertions(+)
> >>>
> >>> diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
> >>> index 543898482f8b..161107f2d3bd 100644
> >>> --- a/fs/f2fs/f2fs.h
> >>> +++ b/fs/f2fs/f2fs.h
> >>> @@ -1558,6 +1558,7 @@ struct f2fs_sb_info {
> >>>
> >>>    #ifdef CONFIG_BLK_DEV_ZONED
> >>>        unsigned int blocks_per_blkz;           /* F2FS blocks per zon=
e */
> >>> +     unsigned int max_open_zones;            /* max open zone resour=
ces of the zoned device */
> >>>    #endif
> >>>
> >>>        /* for node-related operations */
> >>> diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c
> >>> index 1b718bebfaa1..45e82d6016fc 100644
> >>> --- a/fs/f2fs/super.c
> >>> +++ b/fs/f2fs/super.c
> >>> @@ -2388,6 +2388,16 @@ static int f2fs_remount(struct super_block *sb=
, int *flags, char *data)
> >>>        if (err)
> >>>                goto restore_opts;
> >>>
> >>> +#ifdef CONFIG_BLK_DEV_ZONED
> >>> +     if (sbi->max_open_zones && sbi->max_open_zones < F2FS_OPTION(sb=
i).active_logs) {
> >>> +             f2fs_err(sbi,
> >>> +                     "zoned: max open zones %u is too small, need at=
 least %u open zones",
> >>> +                              sbi->max_open_zones, F2FS_OPTION(sbi).=
active_logs);
> >>> +             err =3D -EINVAL;
> >>> +             goto restore_opts;
> >>> +     }
> >>> +#endif
> >>> +
> >>>        /* flush outstanding errors before changing fs state */
> >>>        flush_work(&sbi->s_error_work);
> >>>
> >>> @@ -3930,11 +3940,22 @@ static int init_blkz_info(struct f2fs_sb_info=
 *sbi, int devi)
> >>>        sector_t nr_sectors =3D bdev_nr_sectors(bdev);
> >>>        struct f2fs_report_zones_args rep_zone_arg;
> >>>        u64 zone_sectors;
> >>> +     unsigned int max_open_zones;
> >>>        int ret;
> >>>
> >>>        if (!f2fs_sb_has_blkzoned(sbi))
> >>>                return 0;
> >>>
> >>> +     max_open_zones =3D bdev_max_open_zones(bdev);
> >>
> >> Wenjie,
> >>
> >> max_open_zones can always be zero? then sbi->max_open_zones will be ze=
ro,
> >> is this a valid case?
> >>
> >> Thanks,
> >>
> >>> +     if (max_open_zones && (max_open_zones < sbi->max_open_zones || =
!sbi->max_open_zones))
> >>> +             sbi->max_open_zones =3D max_open_zones;
> >>> +     if (sbi->max_open_zones && sbi->max_open_zones < F2FS_OPTION(sb=
i).active_logs) {
> >>> +             f2fs_err(sbi,
> >>> +                     "zoned: max open zones %u is too small, need at=
 least %u open zones",
> >>> +                              sbi->max_open_zones, F2FS_OPTION(sbi).=
active_logs);
> >>> +             return -EINVAL;
> >>> +     }
> >>> +
> >>>        zone_sectors =3D bdev_zone_sectors(bdev);
> >>>        if (!is_power_of_2(zone_sectors)) {
> >>>                f2fs_err(sbi, "F2FS does not support non power of 2 zo=
ne sizes\n");