diff mbox series

[f2fs-dev,v3] f2fs:Add write priority option based on zone UFS

Message ID 20240709021614.17044-1-liaoyuanhong@vivo.com (mailing list archive)
State Superseded
Headers show
Series [f2fs-dev,v3] f2fs:Add write priority option based on zone UFS | expand

Commit Message

Liao Yuanhong July 9, 2024, 2:16 a.m. UTC
Currently, we are using a mix of traditional UFS and zone UFS to support 
some functionalities that cannot be achieved on zone UFS alone. However, 
there are some issues with this approach. There exists a significant 
performance difference between traditional UFS and zone UFS. Under normal 
usage, we prioritize writes to zone UFS. However, in critical conditions 
(such as when the entire UFS is almost full), we cannot determine whether 
data will be written to traditional UFS or zone UFS. This can lead to 
significant performance fluctuations, which is not conducive to 
development and testing. To address this, we have added an option 
zlu_io_enable under sys with the following three modes:
1) zlu_io_enable == 0:Normal mode, prioritize writing to zone UFS;
2) zlu_io_enable == 1:Zone UFS only mode, only allow writing to zone UFS;
3) zlu_io_enable == 2:Traditional UFS priority mode, prioritize writing to 
traditional UFS.

Signed-off-by: Liao Yuanhong <liaoyuanhong@vivo.com>
Signed-off-by: Wu Bo <bo.wu@vivo.com>
---
v3:
	-Delete excess comments,
	-Use enum instead of define,
	-Modify some judgment criteria to make them more reasonable.
---
v2:
	-Change name to blkzone_alloc_policy,
	-Update manual of f2fs sysfs entry,
	-Use macro instead of magic number,
	-Initialize it w/ default policy in f2fs_scan_devices,
	-Add validation check,
	-Merged the ifdef PROFIG-BLK-DEV_ZONED area.
---
 Documentation/ABI/testing/sysfs-fs-f2fs | 14 ++++++++++++++
 fs/f2fs/f2fs.h                          |  8 ++++++++
 fs/f2fs/segment.c                       | 25 ++++++++++++++++++++++++-
 fs/f2fs/super.c                         |  1 +
 fs/f2fs/sysfs.c                         | 11 +++++++++++
 5 files changed, 58 insertions(+), 1 deletion(-)

Comments

Chao Yu July 9, 2024, 9:44 a.m. UTC | #1
On 2024/7/9 10:16, Liao Yuanhong wrote:
> Currently, we are using a mix of traditional UFS and zone UFS to support
> some functionalities that cannot be achieved on zone UFS alone. However,
> there are some issues with this approach. There exists a significant
> performance difference between traditional UFS and zone UFS. Under normal
> usage, we prioritize writes to zone UFS. However, in critical conditions
> (such as when the entire UFS is almost full), we cannot determine whether
> data will be written to traditional UFS or zone UFS. This can lead to
> significant performance fluctuations, which is not conducive to
> development and testing. To address this, we have added an option
> zlu_io_enable under sys with the following three modes:
> 1) zlu_io_enable == 0:Normal mode, prioritize writing to zone UFS;
> 2) zlu_io_enable == 1:Zone UFS only mode, only allow writing to zone UFS;
> 3) zlu_io_enable == 2:Traditional UFS priority mode, prioritize writing to
> traditional UFS.
> 
> Signed-off-by: Liao Yuanhong <liaoyuanhong@vivo.com>
> Signed-off-by: Wu Bo <bo.wu@vivo.com>
> ---
> v3:
> 	-Delete excess comments,
> 	-Use enum instead of define,
> 	-Modify some judgment criteria to make them more reasonable.
> ---
> v2:
> 	-Change name to blkzone_alloc_policy,
> 	-Update manual of f2fs sysfs entry,
> 	-Use macro instead of magic number,
> 	-Initialize it w/ default policy in f2fs_scan_devices,
> 	-Add validation check,
> 	-Merged the ifdef PROFIG-BLK-DEV_ZONED area.
> ---
>   Documentation/ABI/testing/sysfs-fs-f2fs | 14 ++++++++++++++
>   fs/f2fs/f2fs.h                          |  8 ++++++++
>   fs/f2fs/segment.c                       | 25 ++++++++++++++++++++++++-
>   fs/f2fs/super.c                         |  1 +
>   fs/f2fs/sysfs.c                         | 11 +++++++++++
>   5 files changed, 58 insertions(+), 1 deletion(-)
> 
> diff --git a/Documentation/ABI/testing/sysfs-fs-f2fs b/Documentation/ABI/testing/sysfs-fs-f2fs
> index cad6c3dc1f9c..3500920ab7ce 100644
> --- a/Documentation/ABI/testing/sysfs-fs-f2fs
> +++ b/Documentation/ABI/testing/sysfs-fs-f2fs
> @@ -763,3 +763,17 @@ Date:		November 2023
>   Contact:	"Chao Yu" <chao@kernel.org>
>   Description:	It controls to enable/disable IO aware feature for background discard.
>   		By default, the value is 1 which indicates IO aware is on.
> +
> +What:		/sys/fs/f2fs/<disk>/blkzone_alloc_policy
> +Date:		July 2024
> +Contact:	"Yuanhong Liao" <liaoyuanhong@vivo.com>
> +Description:	The zone UFS we are currently using consists of two parts:
> +		conventional zones and sequential zones. It can be used to control which part
> +		to prioritize for writes, with a default value of 0.
> +
> +		========================  =========================================
> +		value					  description
> +		blkzone_alloc_policy = 0  Prioritize writing to sequential zones
> +		blkzone_alloc_policy = 1  Only allow writing to sequential zones
> +		blkzone_alloc_policy = 2  Prioritize writing to conventional zones
> +		========================  =========================================
> diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
> index f7ee6c5e371e..adefd19810ff 100644
> --- a/fs/f2fs/f2fs.h
> +++ b/fs/f2fs/f2fs.h
> @@ -134,6 +134,12 @@ typedef u32 nid_t;
>   
>   #define COMPRESS_EXT_NUM		16
>   
> +enum blkzone_allocation_policy {
> +	BLKZONE_ALLOC_PRIOR_SEQ;	/* Prioritize writing to sequential zones */
> +	BLKZONE_ALLOC_ONLY_SEQ;		/* Only allow writing to sequential zones */
> +	BLKZONE_ALLOC_PRIOR_CONV;	/* Prioritize writing to conventional zones */
> +};
> +
>   /*
>    * An implementation of an rwsem that is explicitly unfair to readers. This
>    * prevents priority inversion when a low-priority reader acquires the read lock
> @@ -1555,6 +1561,8 @@ 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 */
> +	/* For adjust the priority writing position of data in zone UFS */
> +	unsigned int blkzone_alloc_policy;
>   #endif
>   
>   	/* for node-related operations */
> diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c
> index 4db1add43e36..b38b92e08c6c 100644
> --- a/fs/f2fs/segment.c
> +++ b/fs/f2fs/segment.c
> @@ -2686,17 +2686,40 @@ static int get_new_segment(struct f2fs_sb_info *sbi,
>   			goto got_it;
>   	}
>   
> +#ifdef CONFIG_BLK_DEV_ZONED
>   	/*
>   	 * If we format f2fs on zoned storage, let's try to get pinned sections
>   	 * from beginning of the storage, which should be a conventional one.
>   	 */
>   	if (f2fs_sb_has_blkzoned(sbi)) {
> -		segno = pinning ? 0 : max(first_zoned_segno(sbi), *newseg);
> +		/* Prioritize writing to conventional zones */
> +		if (sbi->blkzone_alloc_policy == PRIOR_CONVENTIONAL || pinning)

should adjust new enum definition?

> +			segno = 0;
> +		else
> +			segno = max(first_zoned_segno(sbi), *newseg);
>   		hint = GET_SEC_FROM_SEG(sbi, segno);
>   	}
> +#endif
>   
>   find_other_zone:
>   	secno = find_next_zero_bit(free_i->free_secmap, MAIN_SECS(sbi), hint);
> +
> +#ifdef CONFIG_BLK_DEV_ZONED
> +	if (secno >= MAIN_SECS(sbi) && f2fs_sb_has_blkzoned(sbi)) {
> +		/* Write only to sequential zones */
> +		if (sbi->blkzone_alloc_policy == ONLY_SEQUENTIAL) {

ditto

> +			hint = GET_SEC_FROM_SEG(sbi, first_zoned_segno(sbi));
> +			secno = find_next_zero_bit(free_i->free_secmap, MAIN_SECS(sbi), hint);
> +		} else
> +			secno = find_first_zero_bit(free_i->free_secmap,
> +								MAIN_SECS(sbi));
> +		if (secno >= MAIN_SECS(sbi)) {
> +			ret = -ENOSPC;
> +			goto out_unlock;
> +		}
> +	}
> +#endif
> +
>   	if (secno >= MAIN_SECS(sbi)) {
>   		secno = find_first_zero_bit(free_i->free_secmap,
>   							MAIN_SECS(sbi));
> diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c
> index 4a1bc8f40f9a..d5b0b7b141ce 100644
> --- a/fs/f2fs/super.c
> +++ b/fs/f2fs/super.c
> @@ -4229,6 +4229,7 @@ static int f2fs_scan_devices(struct f2fs_sb_info *sbi)
>   	sbi->aligned_blksize = true;
>   #ifdef CONFIG_BLK_DEV_ZONED
>   	sbi->max_open_zones = UINT_MAX;
> +	sbi->blkzone_alloc_policy = PRIOR_SEQUENTIAL;

ditto

>   #endif
>   
>   	for (i = 0; i < max_devices; i++) {
> diff --git a/fs/f2fs/sysfs.c b/fs/f2fs/sysfs.c
> index fee7ee45ceaa..359a12f84060 100644
> --- a/fs/f2fs/sysfs.c
> +++ b/fs/f2fs/sysfs.c
> @@ -627,6 +627,15 @@ static ssize_t __sbi_store(struct f2fs_attr *a,
>   	}
>   #endif
>   
> +#ifdef CONFIG_BLK_DEV_ZONED
> +	if (!strcmp(a->attr.name, "blkzone_alloc_policy")) {
> +		if (t < PRIOR_SEQUENTIAL || t > PRIOR_CONVENTIONAL)

ditto

> +			return -EINVAL;
> +		sbi->blkzone_alloc_policy = t;
> +		return count;
> +	}
> +#endif
> +
>   #ifdef CONFIG_F2FS_FS_COMPRESSION
>   	if (!strcmp(a->attr.name, "compr_written_block") ||
>   		!strcmp(a->attr.name, "compr_saved_block")) {
> @@ -1033,6 +1042,7 @@ F2FS_SBI_GENERAL_RW_ATTR(warm_data_age_threshold);
>   F2FS_SBI_GENERAL_RW_ATTR(last_age_weight);
>   #ifdef CONFIG_BLK_DEV_ZONED
>   F2FS_SBI_GENERAL_RO_ATTR(unusable_blocks_per_sec);
> +F2FS_SBI_GENERAL_RW_ATTR(blkzone_alloc_policy);
>   #endif
>   
>   /* STAT_INFO ATTR */
> @@ -1187,6 +1197,7 @@ static struct attribute *f2fs_attrs[] = {
>   #endif
>   #ifdef CONFIG_BLK_DEV_ZONED
>   	ATTR_LIST(unusable_blocks_per_sec),
> +	ATTR_LIST(blkzone_alloc_policy),
>   #endif
>   #ifdef CONFIG_F2FS_FS_COMPRESSION
>   	ATTR_LIST(compr_written_block),
diff mbox series

Patch

diff --git a/Documentation/ABI/testing/sysfs-fs-f2fs b/Documentation/ABI/testing/sysfs-fs-f2fs
index cad6c3dc1f9c..3500920ab7ce 100644
--- a/Documentation/ABI/testing/sysfs-fs-f2fs
+++ b/Documentation/ABI/testing/sysfs-fs-f2fs
@@ -763,3 +763,17 @@  Date:		November 2023
 Contact:	"Chao Yu" <chao@kernel.org>
 Description:	It controls to enable/disable IO aware feature for background discard.
 		By default, the value is 1 which indicates IO aware is on.
+
+What:		/sys/fs/f2fs/<disk>/blkzone_alloc_policy
+Date:		July 2024
+Contact:	"Yuanhong Liao" <liaoyuanhong@vivo.com>
+Description:	The zone UFS we are currently using consists of two parts:
+		conventional zones and sequential zones. It can be used to control which part
+		to prioritize for writes, with a default value of 0.
+
+		========================  =========================================
+		value					  description
+		blkzone_alloc_policy = 0  Prioritize writing to sequential zones
+		blkzone_alloc_policy = 1  Only allow writing to sequential zones
+		blkzone_alloc_policy = 2  Prioritize writing to conventional zones
+		========================  =========================================
diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
index f7ee6c5e371e..adefd19810ff 100644
--- a/fs/f2fs/f2fs.h
+++ b/fs/f2fs/f2fs.h
@@ -134,6 +134,12 @@  typedef u32 nid_t;
 
 #define COMPRESS_EXT_NUM		16
 
+enum blkzone_allocation_policy {
+	BLKZONE_ALLOC_PRIOR_SEQ;	/* Prioritize writing to sequential zones */
+	BLKZONE_ALLOC_ONLY_SEQ;		/* Only allow writing to sequential zones */
+	BLKZONE_ALLOC_PRIOR_CONV;	/* Prioritize writing to conventional zones */
+};
+
 /*
  * An implementation of an rwsem that is explicitly unfair to readers. This
  * prevents priority inversion when a low-priority reader acquires the read lock
@@ -1555,6 +1561,8 @@  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 */
+	/* For adjust the priority writing position of data in zone UFS */
+	unsigned int blkzone_alloc_policy;
 #endif
 
 	/* for node-related operations */
diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c
index 4db1add43e36..b38b92e08c6c 100644
--- a/fs/f2fs/segment.c
+++ b/fs/f2fs/segment.c
@@ -2686,17 +2686,40 @@  static int get_new_segment(struct f2fs_sb_info *sbi,
 			goto got_it;
 	}
 
+#ifdef CONFIG_BLK_DEV_ZONED
 	/*
 	 * If we format f2fs on zoned storage, let's try to get pinned sections
 	 * from beginning of the storage, which should be a conventional one.
 	 */
 	if (f2fs_sb_has_blkzoned(sbi)) {
-		segno = pinning ? 0 : max(first_zoned_segno(sbi), *newseg);
+		/* Prioritize writing to conventional zones */
+		if (sbi->blkzone_alloc_policy == PRIOR_CONVENTIONAL || pinning)
+			segno = 0;
+		else
+			segno = max(first_zoned_segno(sbi), *newseg);
 		hint = GET_SEC_FROM_SEG(sbi, segno);
 	}
+#endif
 
 find_other_zone:
 	secno = find_next_zero_bit(free_i->free_secmap, MAIN_SECS(sbi), hint);
+
+#ifdef CONFIG_BLK_DEV_ZONED
+	if (secno >= MAIN_SECS(sbi) && f2fs_sb_has_blkzoned(sbi)) {
+		/* Write only to sequential zones */
+		if (sbi->blkzone_alloc_policy == ONLY_SEQUENTIAL) {
+			hint = GET_SEC_FROM_SEG(sbi, first_zoned_segno(sbi));
+			secno = find_next_zero_bit(free_i->free_secmap, MAIN_SECS(sbi), hint);
+		} else
+			secno = find_first_zero_bit(free_i->free_secmap,
+								MAIN_SECS(sbi));
+		if (secno >= MAIN_SECS(sbi)) {
+			ret = -ENOSPC;
+			goto out_unlock;
+		}
+	}
+#endif
+
 	if (secno >= MAIN_SECS(sbi)) {
 		secno = find_first_zero_bit(free_i->free_secmap,
 							MAIN_SECS(sbi));
diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c
index 4a1bc8f40f9a..d5b0b7b141ce 100644
--- a/fs/f2fs/super.c
+++ b/fs/f2fs/super.c
@@ -4229,6 +4229,7 @@  static int f2fs_scan_devices(struct f2fs_sb_info *sbi)
 	sbi->aligned_blksize = true;
 #ifdef CONFIG_BLK_DEV_ZONED
 	sbi->max_open_zones = UINT_MAX;
+	sbi->blkzone_alloc_policy = PRIOR_SEQUENTIAL;
 #endif
 
 	for (i = 0; i < max_devices; i++) {
diff --git a/fs/f2fs/sysfs.c b/fs/f2fs/sysfs.c
index fee7ee45ceaa..359a12f84060 100644
--- a/fs/f2fs/sysfs.c
+++ b/fs/f2fs/sysfs.c
@@ -627,6 +627,15 @@  static ssize_t __sbi_store(struct f2fs_attr *a,
 	}
 #endif
 
+#ifdef CONFIG_BLK_DEV_ZONED
+	if (!strcmp(a->attr.name, "blkzone_alloc_policy")) {
+		if (t < PRIOR_SEQUENTIAL || t > PRIOR_CONVENTIONAL)
+			return -EINVAL;
+		sbi->blkzone_alloc_policy = t;
+		return count;
+	}
+#endif
+
 #ifdef CONFIG_F2FS_FS_COMPRESSION
 	if (!strcmp(a->attr.name, "compr_written_block") ||
 		!strcmp(a->attr.name, "compr_saved_block")) {
@@ -1033,6 +1042,7 @@  F2FS_SBI_GENERAL_RW_ATTR(warm_data_age_threshold);
 F2FS_SBI_GENERAL_RW_ATTR(last_age_weight);
 #ifdef CONFIG_BLK_DEV_ZONED
 F2FS_SBI_GENERAL_RO_ATTR(unusable_blocks_per_sec);
+F2FS_SBI_GENERAL_RW_ATTR(blkzone_alloc_policy);
 #endif
 
 /* STAT_INFO ATTR */
@@ -1187,6 +1197,7 @@  static struct attribute *f2fs_attrs[] = {
 #endif
 #ifdef CONFIG_BLK_DEV_ZONED
 	ATTR_LIST(unusable_blocks_per_sec),
+	ATTR_LIST(blkzone_alloc_policy),
 #endif
 #ifdef CONFIG_F2FS_FS_COMPRESSION
 	ATTR_LIST(compr_written_block),