diff mbox series

[v4] f2fs: fix zoned block device information initialization

Message ID 20240203152436.1352-1-qwjhust@gmail.com
State New
Headers
Series [v4] f2fs: fix zoned block device information initialization |

Commit Message

Wenjie Qi Feb. 3, 2024, 3:24 p.m. UTC 
  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(+)

Comments

Chao Yu Feb. 4, 2024, 1:47 a.m. UTC | #1 
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 zone */
> +	unsigned int max_open_zones;		/* max open zone resources 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(sbi).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 = -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 = 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 = bdev_max_open_zones(bdev);

Wenjie,

max_open_zones can always be zero? then sbi->max_open_zones will be zero,
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 = max_open_zones;
> +	if (sbi->max_open_zones && sbi->max_open_zones < F2FS_OPTION(sbi).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 = bdev_zone_sectors(bdev);
>   	if (!is_power_of_2(zone_sectors)) {
>   		f2fs_err(sbi, "F2FS does not support non power of 2 zone sizes\n");
Wenjie Qi Feb. 4, 2024, 2:18 a.m. UTC | #2 
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.

Thanks,

Chao Yu <chao@kernel.org> 于2024年2月4日周日 09:47写道：
>
> 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 zone */
> > +     unsigned int max_open_zones;            /* max open zone resources 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(sbi)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 = -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 = 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 = bdev_max_open_zones(bdev);
>
> Wenjie,
>
> max_open_zones can always be zero? then sbi->max_open_zones will be zero,
> 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 = max_open_zones;
> > +     if (sbi->max_open_zones && sbi->max_open_zones < F2FS_OPTION(sbi)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 = bdev_zone_sectors(bdev);
> >       if (!is_power_of_2(zone_sectors)) {
> >               f2fs_err(sbi, "F2FS does not support non power of 2 zone sizes\n");
Chao Yu Feb. 4, 2024, 2:24 a.m. UTC | #3 
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> 于2024年2月4日周日 09:47写道：
>>
>> 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 zone */
>>> +     unsigned int max_open_zones;            /* max open zone resources 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(sbi).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 = -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 = 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 = bdev_max_open_zones(bdev);
>>
>> Wenjie,
>>
>> max_open_zones can always be zero? then sbi->max_open_zones will be zero,
>> 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 = max_open_zones;
>>> +     if (sbi->max_open_zones && sbi->max_open_zones < F2FS_OPTION(sbi).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 = bdev_zone_sectors(bdev);
>>>        if (!is_power_of_2(zone_sectors)) {
>>>                f2fs_err(sbi, "F2FS does not support non power of 2 zone sizes\n");
Wenjie Qi Feb. 4, 2024, 2:36 a.m. UTC | #4 
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 AM 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> 于2024年2月4日周日 09:47写道：
> >>
> >> 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 zone */
> >>> +     unsigned int max_open_zones;            /* max open zone resources 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(sbi).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 = -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 = 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 = bdev_max_open_zones(bdev);
> >>
> >> Wenjie,
> >>
> >> max_open_zones can always be zero? then sbi->max_open_zones will be zero,
> >> 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 = max_open_zones;
> >>> +     if (sbi->max_open_zones && sbi->max_open_zones < F2FS_OPTION(sbi).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 = bdev_zone_sectors(bdev);
> >>>        if (!is_power_of_2(zone_sectors)) {
> >>>                f2fs_err(sbi, "F2FS does not support non power of 2 zone sizes\n");
diff mbox series

Patch

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 zone */
+	unsigned int max_open_zones;		/* max open zone resources 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(sbi).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 = -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 = 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 = bdev_max_open_zones(bdev);
+	if (max_open_zones && (max_open_zones < sbi->max_open_zones || !sbi->max_open_zones))
+		sbi->max_open_zones = max_open_zones;
+	if (sbi->max_open_zones && sbi->max_open_zones < F2FS_OPTION(sbi).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 = bdev_zone_sectors(bdev);
 	if (!is_power_of_2(zone_sectors)) {
 		f2fs_err(sbi, "F2FS does not support non power of 2 zone sizes\n");