Export max open zones and max active zones to sysfs.
This patch series in based on Jens's linux-block/for-next branch.
All zoned block devices in the kernel utilize the "zoned block device
support" (CONFIG_BLK_DEV_ZONED).
The Zoned Namespace Command Set Specification defines two different
resource limits: Max Open Resources and Max Active Resources.
The ZAC and ZBC standards define a MAXIMUM NUMBER OF OPEN SEQUENTIAL WRITE
REQUIRED ZONES field.
Since the ZNS Max Open Resources field has the same purpose as the ZAC/ZBC
field, (the ZNS field is 0's based, the ZAC/ZBC field isn't), create a
common "max_open_zones" definition in the sysfs documentation, and export
both the ZNS field and the ZAC/ZBC field according to this new common
definition.
The ZNS Max Active Resources field does not have an equivalent field in
ZAC/ZBC, however, since both ZAC/ZBC and ZNS utilize the "zoned block
device support" in the kernel, create a "max_active_zones" definition in
the sysfs documentation, similar to "max_open_zones", and export it
according to this new definition. For ZAC/ZBC devices, this field will be
exported as 0, meaning "no limit".
Changes since v2:
-Picked up Damien's Reviewed-by tags.
-Update Documentation/ABI/testing/sysfs-block in addition to
Documentation/block/queue-sysfs.rst (Greg).
-Added bdev_max_open_zones()/bdev_max_active_zones() helpers
(Johannes).
Niklas Cassel (2):
block: add max_open_zones to blk-sysfs
block: add max_active_zones to blk-sysfs
Documentation/ABI/testing/sysfs-block | 18 ++++++++++
Documentation/block/queue-sysfs.rst | 14 ++++++++
block/blk-sysfs.c | 27 +++++++++++++++
drivers/nvme/host/zns.c | 2 ++
drivers/scsi/sd_zbc.c | 5 +++
include/linux/blkdev.h | 50 +++++++++++++++++++++++++++
6 files changed, 116 insertions(+)
--
2.26.2
Add a new max_active zones definition in the sysfs documentation.
This definition will be common for all devices utilizing the zoned block
device support in the kernel.
Export max_active_zones according to this new definition for NVMe Zoned
Namespace devices, ZAC ATA devices (which are treated as SCSI devices by
the kernel), and ZBC SCSI devices.
Add the new max_active_zones member to struct request_queue, rather
than as a queue limit, since this property cannot be split across stacking
drivers.
For SCSI devices, even though max active zones is not part of the ZBC/ZAC
spec, export max_active_zones as 0, signifying "no limit".
Signed-off-by: Niklas Cassel <[email protected]>
Reviewed-by: Javier González <[email protected]>
Reviewed-by: Damien Le Moal <[email protected]>
---
Documentation/ABI/testing/sysfs-block | 9 +++++++++
Documentation/block/queue-sysfs.rst | 7 +++++++
block/blk-sysfs.c | 14 +++++++++++++-
drivers/nvme/host/zns.c | 1 +
drivers/scsi/sd_zbc.c | 1 +
include/linux/blkdev.h | 25 +++++++++++++++++++++++++
6 files changed, 56 insertions(+), 1 deletion(-)
diff --git a/Documentation/ABI/testing/sysfs-block b/Documentation/ABI/testing/sysfs-block
index f151d9cf90de..2322eb748b38 100644
--- a/Documentation/ABI/testing/sysfs-block
+++ b/Documentation/ABI/testing/sysfs-block
@@ -273,6 +273,15 @@ Description:
device ("host-aware" or "host-managed" zone model). For regular
block devices, the value is always 0.
+What: /sys/block/<disk>/queue/max_active_zones
+Date: July 2020
+Contact: Niklas Cassel <[email protected]>
+Description:
+ For zoned block devices (zoned attribute indicating
+ "host-managed" or "host-aware"), the sum of zones belonging to
+ any of the zone states: EXPLICIT OPEN, IMPLICIT OPEN or CLOSED,
+ is limited by this value. If this value is 0, there is no limit.
+
What: /sys/block/<disk>/queue/max_open_zones
Date: July 2020
Contact: Niklas Cassel <[email protected]>
diff --git a/Documentation/block/queue-sysfs.rst b/Documentation/block/queue-sysfs.rst
index f01cf8530ae4..f261a5c84170 100644
--- a/Documentation/block/queue-sysfs.rst
+++ b/Documentation/block/queue-sysfs.rst
@@ -117,6 +117,13 @@ Maximum number of elements in a DMA scatter/gather list with integrity
data that will be submitted by the block layer core to the associated
block driver.
+max_active_zones (RO)
+---------------------
+For zoned block devices (zoned attribute indicating "host-managed" or
+"host-aware"), the sum of zones belonging to any of the zone states:
+EXPLICIT OPEN, IMPLICIT OPEN or CLOSED, is limited by this value.
+If this value is 0, there is no limit.
+
max_open_zones (RO)
-------------------
For zoned block devices (zoned attribute indicating "host-managed" or
diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c
index 414f04579d77..7dda709f3ccb 100644
--- a/block/blk-sysfs.c
+++ b/block/blk-sysfs.c
@@ -311,6 +311,11 @@ static ssize_t queue_max_open_zones_show(struct request_queue *q, char *page)
return queue_var_show(queue_max_open_zones(q), page);
}
+static ssize_t queue_max_active_zones_show(struct request_queue *q, char *page)
+{
+ return queue_var_show(queue_max_active_zones(q), page);
+}
+
static ssize_t queue_nomerges_show(struct request_queue *q, char *page)
{
return queue_var_show((blk_queue_nomerges(q) << 1) |
@@ -678,6 +683,11 @@ static struct queue_sysfs_entry queue_max_open_zones_entry = {
.show = queue_max_open_zones_show,
};
+static struct queue_sysfs_entry queue_max_active_zones_entry = {
+ .attr = {.name = "max_active_zones", .mode = 0444 },
+ .show = queue_max_active_zones_show,
+};
+
static struct queue_sysfs_entry queue_nomerges_entry = {
.attr = {.name = "nomerges", .mode = 0644 },
.show = queue_nomerges_show,
@@ -777,6 +787,7 @@ static struct attribute *queue_attrs[] = {
&queue_zoned_entry.attr,
&queue_nr_zones_entry.attr,
&queue_max_open_zones_entry.attr,
+ &queue_max_active_zones_entry.attr,
&queue_nomerges_entry.attr,
&queue_rq_affinity_entry.attr,
&queue_iostats_entry.attr,
@@ -804,7 +815,8 @@ static umode_t queue_attr_visible(struct kobject *kobj, struct attribute *attr,
(!q->mq_ops || !q->mq_ops->timeout))
return 0;
- if (attr == &queue_max_open_zones_entry.attr &&
+ if ((attr == &queue_max_open_zones_entry.attr ||
+ attr == &queue_max_active_zones_entry.attr) &&
!blk_queue_is_zoned(q))
return 0;
diff --git a/drivers/nvme/host/zns.c b/drivers/nvme/host/zns.c
index 3d80b9cf6bfc..57cfd78731fb 100644
--- a/drivers/nvme/host/zns.c
+++ b/drivers/nvme/host/zns.c
@@ -97,6 +97,7 @@ int nvme_update_zone_info(struct gendisk *disk, struct nvme_ns *ns,
q->limits.zoned = BLK_ZONED_HM;
blk_queue_flag_set(QUEUE_FLAG_ZONE_RESETALL, q);
blk_queue_max_open_zones(q, le32_to_cpu(id->mor) + 1);
+ blk_queue_max_active_zones(q, le32_to_cpu(id->mar) + 1);
free_data:
kfree(id);
return status;
diff --git a/drivers/scsi/sd_zbc.c b/drivers/scsi/sd_zbc.c
index aa3564139b40..d8b2c49d645b 100644
--- a/drivers/scsi/sd_zbc.c
+++ b/drivers/scsi/sd_zbc.c
@@ -721,6 +721,7 @@ int sd_zbc_read_zones(struct scsi_disk *sdkp, unsigned char *buf)
blk_queue_max_open_zones(q, 0);
else
blk_queue_max_open_zones(q, sdkp->zones_max_open);
+ blk_queue_max_active_zones(q, 0);
nr_zones = round_up(sdkp->capacity, zone_blocks) >> ilog2(zone_blocks);
/* READ16/WRITE16 is mandatory for ZBC disks */
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index 8f558c6fd18b..692ddaf07dc4 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -514,6 +514,7 @@ struct request_queue {
unsigned long *conv_zones_bitmap;
unsigned long *seq_zones_wlock;
unsigned int max_open_zones;
+ unsigned int max_active_zones;
#endif /* CONFIG_BLK_DEV_ZONED */
/*
@@ -734,6 +735,17 @@ static inline unsigned int queue_max_open_zones(const struct request_queue *q)
{
return q->max_open_zones;
}
+
+static inline void blk_queue_max_active_zones(struct request_queue *q,
+ unsigned int max_active_zones)
+{
+ q->max_active_zones = max_active_zones;
+}
+
+static inline unsigned int queue_max_active_zones(const struct request_queue *q)
+{
+ return q->max_active_zones;
+}
#else /* CONFIG_BLK_DEV_ZONED */
static inline unsigned int blk_queue_nr_zones(struct request_queue *q)
{
@@ -753,6 +765,10 @@ static inline unsigned int queue_max_open_zones(const struct request_queue *q)
{
return 0;
}
+static inline unsigned int queue_max_active_zones(const struct request_queue *q)
+{
+ return 0;
+}
#endif /* CONFIG_BLK_DEV_ZONED */
static inline bool rq_is_sync(struct request *rq)
@@ -1545,6 +1561,15 @@ static inline unsigned int bdev_max_open_zones(struct block_device *bdev)
return 0;
}
+static inline unsigned int bdev_max_active_zones(struct block_device *bdev)
+{
+ struct request_queue *q = bdev_get_queue(bdev);
+
+ if (q)
+ return queue_max_active_zones(q);
+ return 0;
+}
+
static inline int queue_dma_alignment(const struct request_queue *q)
{
return q ? q->dma_alignment : 511;
--
2.26.2
Add a new max_open_zones definition in the sysfs documentation.
This definition will be common for all devices utilizing the zoned block
device support in the kernel.
Export max open zones according to this new definition for NVMe Zoned
Namespace devices, ZAC ATA devices (which are treated as SCSI devices by
the kernel), and ZBC SCSI devices.
Add the new max_open_zones member to struct request_queue, rather
than as a queue limit, since this property cannot be split across stacking
drivers.
Signed-off-by: Niklas Cassel <[email protected]>
Reviewed-by: Javier González <[email protected]>
Reviewed-by: Damien Le Moal <[email protected]>
---
Documentation/ABI/testing/sysfs-block | 9 +++++++++
Documentation/block/queue-sysfs.rst | 7 +++++++
block/blk-sysfs.c | 15 +++++++++++++++
drivers/nvme/host/zns.c | 1 +
drivers/scsi/sd_zbc.c | 4 ++++
include/linux/blkdev.h | 25 +++++++++++++++++++++++++
6 files changed, 61 insertions(+)
diff --git a/Documentation/ABI/testing/sysfs-block b/Documentation/ABI/testing/sysfs-block
index ed8c14f161ee..f151d9cf90de 100644
--- a/Documentation/ABI/testing/sysfs-block
+++ b/Documentation/ABI/testing/sysfs-block
@@ -273,6 +273,15 @@ Description:
device ("host-aware" or "host-managed" zone model). For regular
block devices, the value is always 0.
+What: /sys/block/<disk>/queue/max_open_zones
+Date: July 2020
+Contact: Niklas Cassel <[email protected]>
+Description:
+ For zoned block devices (zoned attribute indicating
+ "host-managed" or "host-aware"), the sum of zones belonging to
+ any of the zone states: EXPLICIT OPEN or IMPLICIT OPEN,
+ is limited by this value. If this value is 0, there is no limit.
+
What: /sys/block/<disk>/queue/chunk_sectors
Date: September 2016
Contact: Hannes Reinecke <[email protected]>
diff --git a/Documentation/block/queue-sysfs.rst b/Documentation/block/queue-sysfs.rst
index 6a8513af9201..f01cf8530ae4 100644
--- a/Documentation/block/queue-sysfs.rst
+++ b/Documentation/block/queue-sysfs.rst
@@ -117,6 +117,13 @@ Maximum number of elements in a DMA scatter/gather list with integrity
data that will be submitted by the block layer core to the associated
block driver.
+max_open_zones (RO)
+-------------------
+For zoned block devices (zoned attribute indicating "host-managed" or
+"host-aware"), the sum of zones belonging to any of the zone states:
+EXPLICIT OPEN or IMPLICIT OPEN, is limited by this value.
+If this value is 0, there is no limit.
+
max_sectors_kb (RW)
-------------------
This is the maximum number of kilobytes that the block layer will allow
diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c
index be67952e7be2..414f04579d77 100644
--- a/block/blk-sysfs.c
+++ b/block/blk-sysfs.c
@@ -306,6 +306,11 @@ static ssize_t queue_nr_zones_show(struct request_queue *q, char *page)
return queue_var_show(blk_queue_nr_zones(q), page);
}
+static ssize_t queue_max_open_zones_show(struct request_queue *q, char *page)
+{
+ return queue_var_show(queue_max_open_zones(q), page);
+}
+
static ssize_t queue_nomerges_show(struct request_queue *q, char *page)
{
return queue_var_show((blk_queue_nomerges(q) << 1) |
@@ -668,6 +673,11 @@ static struct queue_sysfs_entry queue_nr_zones_entry = {
.show = queue_nr_zones_show,
};
+static struct queue_sysfs_entry queue_max_open_zones_entry = {
+ .attr = {.name = "max_open_zones", .mode = 0444 },
+ .show = queue_max_open_zones_show,
+};
+
static struct queue_sysfs_entry queue_nomerges_entry = {
.attr = {.name = "nomerges", .mode = 0644 },
.show = queue_nomerges_show,
@@ -766,6 +776,7 @@ static struct attribute *queue_attrs[] = {
&queue_nonrot_entry.attr,
&queue_zoned_entry.attr,
&queue_nr_zones_entry.attr,
+ &queue_max_open_zones_entry.attr,
&queue_nomerges_entry.attr,
&queue_rq_affinity_entry.attr,
&queue_iostats_entry.attr,
@@ -793,6 +804,10 @@ static umode_t queue_attr_visible(struct kobject *kobj, struct attribute *attr,
(!q->mq_ops || !q->mq_ops->timeout))
return 0;
+ if (attr == &queue_max_open_zones_entry.attr &&
+ !blk_queue_is_zoned(q))
+ return 0;
+
return attr->mode;
}
diff --git a/drivers/nvme/host/zns.c b/drivers/nvme/host/zns.c
index 04e5b991c00c..3d80b9cf6bfc 100644
--- a/drivers/nvme/host/zns.c
+++ b/drivers/nvme/host/zns.c
@@ -96,6 +96,7 @@ int nvme_update_zone_info(struct gendisk *disk, struct nvme_ns *ns,
q->limits.zoned = BLK_ZONED_HM;
blk_queue_flag_set(QUEUE_FLAG_ZONE_RESETALL, q);
+ blk_queue_max_open_zones(q, le32_to_cpu(id->mor) + 1);
free_data:
kfree(id);
return status;
diff --git a/drivers/scsi/sd_zbc.c b/drivers/scsi/sd_zbc.c
index 183a20720da9..aa3564139b40 100644
--- a/drivers/scsi/sd_zbc.c
+++ b/drivers/scsi/sd_zbc.c
@@ -717,6 +717,10 @@ int sd_zbc_read_zones(struct scsi_disk *sdkp, unsigned char *buf)
/* The drive satisfies the kernel restrictions: set it up */
blk_queue_flag_set(QUEUE_FLAG_ZONE_RESETALL, q);
blk_queue_required_elevator_features(q, ELEVATOR_F_ZBD_SEQ_WRITE);
+ if (sdkp->zones_max_open == U32_MAX)
+ blk_queue_max_open_zones(q, 0);
+ else
+ blk_queue_max_open_zones(q, sdkp->zones_max_open);
nr_zones = round_up(sdkp->capacity, zone_blocks) >> ilog2(zone_blocks);
/* READ16/WRITE16 is mandatory for ZBC disks */
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index 69ad13dacd48..8f558c6fd18b 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -513,6 +513,7 @@ struct request_queue {
unsigned int nr_zones;
unsigned long *conv_zones_bitmap;
unsigned long *seq_zones_wlock;
+ unsigned int max_open_zones;
#endif /* CONFIG_BLK_DEV_ZONED */
/*
@@ -722,6 +723,17 @@ static inline bool blk_queue_zone_is_seq(struct request_queue *q,
return true;
return !test_bit(blk_queue_zone_no(q, sector), q->conv_zones_bitmap);
}
+
+static inline void blk_queue_max_open_zones(struct request_queue *q,
+ unsigned int max_open_zones)
+{
+ q->max_open_zones = max_open_zones;
+}
+
+static inline unsigned int queue_max_open_zones(const struct request_queue *q)
+{
+ return q->max_open_zones;
+}
#else /* CONFIG_BLK_DEV_ZONED */
static inline unsigned int blk_queue_nr_zones(struct request_queue *q)
{
@@ -737,6 +749,10 @@ static inline unsigned int blk_queue_zone_no(struct request_queue *q,
{
return 0;
}
+static inline unsigned int queue_max_open_zones(const struct request_queue *q)
+{
+ return 0;
+}
#endif /* CONFIG_BLK_DEV_ZONED */
static inline bool rq_is_sync(struct request *rq)
@@ -1520,6 +1536,15 @@ static inline sector_t bdev_zone_sectors(struct block_device *bdev)
return 0;
}
+static inline unsigned int bdev_max_open_zones(struct block_device *bdev)
+{
+ struct request_queue *q = bdev_get_queue(bdev);
+
+ if (q)
+ return queue_max_open_zones(q);
+ return 0;
+}
+
static inline int queue_dma_alignment(const struct request_queue *q)
{
return q ? q->dma_alignment : 511;
--
2.26.2
Looks good,
Reviewed-by: Johannes Thumshirn <[email protected]>
Looks good,
Reviewed-by: Johannes Thumshirn <[email protected]>
Niklas,
> Export max open zones and max active zones to sysfs.
Looks fine.
Reviewed-by: Martin K. Petersen <[email protected]>
--
Martin K. Petersen Oracle Linux Engineering
On 7/14/20 3:18 PM, Niklas Cassel wrote:
> Export max open zones and max active zones to sysfs.
>
> This patch series in based on Jens's linux-block/for-next branch.
>
> All zoned block devices in the kernel utilize the "zoned block device
> support" (CONFIG_BLK_DEV_ZONED).
>
> The Zoned Namespace Command Set Specification defines two different
> resource limits: Max Open Resources and Max Active Resources.
>
> The ZAC and ZBC standards define a MAXIMUM NUMBER OF OPEN SEQUENTIAL WRITE
> REQUIRED ZONES field.
>
>
> Since the ZNS Max Open Resources field has the same purpose as the ZAC/ZBC
> field, (the ZNS field is 0's based, the ZAC/ZBC field isn't), create a
> common "max_open_zones" definition in the sysfs documentation, and export
> both the ZNS field and the ZAC/ZBC field according to this new common
> definition.
>
> The ZNS Max Active Resources field does not have an equivalent field in
> ZAC/ZBC, however, since both ZAC/ZBC and ZNS utilize the "zoned block
> device support" in the kernel, create a "max_active_zones" definition in
> the sysfs documentation, similar to "max_open_zones", and export it
> according to this new definition. For ZAC/ZBC devices, this field will be
> exported as 0, meaning "no limit".
Applied, thanks.
--
Jens Axboe