[PATCH v3 0/3] block: add zone write granularity limit

[PATCH v3 0/3] block: add zone write granularity limit

[Damien Le Moal]

The first patch in this series introduces the zone write granularity
queue limit to indicate the alignment constraint for write operations
into sequential zones of zoned block devices.

The second patch fixes adds the missing documentation for
zone_append_max_bytes to the sysfs block documentation.

The third patch switch zonefs to use this new limit as the file block
size instead of using the physical block size.

Changes from v2:
* Added patch 3 for zonefs
* Addressed Christoph's comments on patch 1 and added the limit
  initialization for zoned nullblk

Changes from v1:
* Fixed typo in patch 2

Damien Le Moal (3):
  block: introduce zone_write_granularity limit
  block: document zone_append_max_bytes attribute
  zonefs: use zone write granularity as block size

 Documentation/block/queue-sysfs.rst | 13 ++++++++++
 block/blk-settings.c                | 39 +++++++++++++++++++++++++----
 block/blk-sysfs.c                   |  8 ++++++
 drivers/block/null_blk/zoned.c      |  1 +
 drivers/nvme/host/zns.c             |  1 +
 drivers/scsi/sd_zbc.c               | 10 ++++++++
 fs/zonefs/super.c                   |  9 +++----
 include/linux/blkdev.h              | 15 +++++++++++
 8 files changed, 86 insertions(+), 10 deletions(-)

-- 
2.29.2

[PATCH v3 0/3] block: add zone write granularity limit

[Damien Le Moal]

The first patch in this series introduces the zone write granularity
queue limit to indicate the alignment constraint for write operations
into sequential zones of zoned block devices.

The second patch fixes adds the missing documentation for
zone_append_max_bytes to the sysfs block documentation.

The third patch switch zonefs to use this new limit as the file block
size instead of using the physical block size.

Changes from v2:
* Added patch 3 for zonefs
* Addressed Christoph's comments on patch 1 and added the limit
  initialization for zoned nullblk

Changes from v1:
* Fixed typo in patch 2

Damien Le Moal (3):
  block: introduce zone_write_granularity limit
  block: document zone_append_max_bytes attribute
  zonefs: use zone write granularity as block size

 Documentation/block/queue-sysfs.rst | 13 ++++++++++
 block/blk-settings.c                | 39 +++++++++++++++++++++++++----
 block/blk-sysfs.c                   |  8 ++++++
 drivers/block/null_blk/zoned.c      |  1 +
 drivers/nvme/host/zns.c             |  1 +
 drivers/scsi/sd_zbc.c               | 10 ++++++++
 fs/zonefs/super.c                   |  9 +++----
 include/linux/blkdev.h              | 15 +++++++++++
 8 files changed, 86 insertions(+), 10 deletions(-)

-- 
2.29.2


_______________________________________________
Linux-nvme mailing list
Linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme

[PATCH v3 1/3] block: introduce zone_write_granularity limit

[Damien Le Moal]

Per ZBC and ZAC specifications, host-managed SMR hard-disks mandate that
all writes into sequential write required zones be aligned to the device
physical block size. However, NVMe ZNS does not have this constraint and
allows write operations into sequential zones to be logical block size
aligned. This inconsistency does not help with portability of software
across device types.
To solve this, introduce the zone_write_granularity queue limit to
indicate the alignment constraint, in bytes, of write operations into
zones of a zoned block device. This new limit is exported as a
read-only sysfs queue attribute and the helper
blk_queue_zone_write_granularity() introduced for drivers to set this
limit. The scsi disk driver is modified to use this helper to set
host-managed SMR disk zone write granularity to the disk physical block
size. The ZNS support code of the NVMe driver is also modified to use
this helper to set the new limit to the logical block size of the
namespace. The nullblk driver is similarly modified too.

The accessor functions queue_zone_write_granularity() and
bdev_zone_write_granularity() are also introduced.

Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>
---
 Documentation/block/queue-sysfs.rst |  7 ++++++
 block/blk-settings.c                | 39 +++++++++++++++++++++++++----
 block/blk-sysfs.c                   |  8 ++++++
 drivers/block/null_blk/zoned.c      |  1 +
 drivers/nvme/host/zns.c             |  1 +
 drivers/scsi/sd_zbc.c               | 10 ++++++++
 include/linux/blkdev.h              | 15 +++++++++++
 7 files changed, 76 insertions(+), 5 deletions(-)

diff --git a/Documentation/block/queue-sysfs.rst b/Documentation/block/queue-sysfs.rst
index 2638d3446b79..c8bf8bc3c03a 100644
--- a/Documentation/block/queue-sysfs.rst
+++ b/Documentation/block/queue-sysfs.rst
@@ -273,4 +273,11 @@ devices are described in the ZBC (Zoned Block Commands) and ZAC
 do not support zone commands, they will be treated as regular block devices
 and zoned will report "none".
 
+zone_write_granularity (RO)
+---------------------------
+This indicates the alignment constraint, in bytes, for write operations in
+sequential zones of zoned block devices (devices with a zoned attributed
+that reports "host-managed" or "host-aware"). This value is always 0 for
+regular block devices.
+
 Jens Axboe <jens.axboe@oracle.com>, February 2009
diff --git a/block/blk-settings.c b/block/blk-settings.c
index 43990b1d148b..48872b4085d4 100644
--- a/block/blk-settings.c
+++ b/block/blk-settings.c
@@ -60,6 +60,7 @@ void blk_set_default_limits(struct queue_limits *lim)
 	lim->io_opt = 0;
 	lim->misaligned = 0;
 	lim->zoned = BLK_ZONED_NONE;
+	lim->zone_write_granularity = 0;
 }
 EXPORT_SYMBOL(blk_set_default_limits);
 
@@ -366,6 +367,28 @@ void blk_queue_physical_block_size(struct request_queue *q, unsigned int size)
 }
 EXPORT_SYMBOL(blk_queue_physical_block_size);
 
+/**
+ * blk_queue_zone_write_granularity - set zone write granularity for the queue
+ * @q:  the request queue for the zoned device
+ * @size:  the zone write granularity size, in bytes
+ *
+ * Description:
+ *   This should be set to the lowest possible size allowing to write in
+ *   sequential zones of a zoned block device.
+ */
+void blk_queue_zone_write_granularity(struct request_queue *q,
+				      unsigned int size)
+{
+	if (WARN_ON_ONCE(!blk_queue_is_zoned(q)))
+		return;
+
+	q->limits.zone_write_granularity = size;
+
+	if (q->limits.zone_write_granularity < q->limits.logical_block_size)
+		q->limits.zone_write_granularity = q->limits.logical_block_size;
+}
+EXPORT_SYMBOL_GPL(blk_queue_zone_write_granularity);
+
 /**
  * blk_queue_alignment_offset - set physical block alignment offset
  * @q:	the request queue for the device
@@ -631,6 +654,8 @@ int blk_stack_limits(struct queue_limits *t, struct queue_limits *b,
 			t->discard_granularity;
 	}
 
+	t->zone_write_granularity = max(t->zone_write_granularity,
+					b->zone_write_granularity);
 	t->zoned = max(t->zoned, b->zoned);
 	return ret;
 }
@@ -847,6 +872,8 @@ EXPORT_SYMBOL_GPL(blk_queue_can_use_dma_map_merging);
  */
 void blk_queue_set_zoned(struct gendisk *disk, enum blk_zoned_model model)
 {
+	struct request_queue *q = disk->queue;
+
 	switch (model) {
 	case BLK_ZONED_HM:
 		/*
@@ -864,18 +891,20 @@ void blk_queue_set_zoned(struct gendisk *disk, enum blk_zoned_model model)
 		 * partitions and zoned block device support is enabled, else
 		 * we do nothing special as far as the block layer is concerned.
 		 */
-		if (!IS_ENABLED(CONFIG_BLK_DEV_ZONED) ||
-		    disk_has_partitions(disk))
-			model = BLK_ZONED_NONE;
-		break;
+		if (IS_ENABLED(CONFIG_BLK_DEV_ZONED) &&
+		    !disk_has_partitions(disk))
+			break;
+		model = BLK_ZONED_NONE;
+		fallthrough;
 	case BLK_ZONED_NONE:
 	default:
 		if (WARN_ON_ONCE(model != BLK_ZONED_NONE))
 			model = BLK_ZONED_NONE;
+		q->limits.zone_write_granularity = 0;
 		break;
 	}
 
-	disk->queue->limits.zoned = model;
+	q->limits.zoned = model;
 }
 EXPORT_SYMBOL_GPL(blk_queue_set_zoned);
 
diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c
index b513f1683af0..ae39c7f3d83d 100644
--- a/block/blk-sysfs.c
+++ b/block/blk-sysfs.c
@@ -219,6 +219,12 @@ static ssize_t queue_write_zeroes_max_show(struct request_queue *q, char *page)
 		(unsigned long long)q->limits.max_write_zeroes_sectors << 9);
 }
 
+static ssize_t queue_zone_write_granularity_show(struct request_queue *q,
+						 char *page)
+{
+	return queue_var_show(queue_zone_write_granularity(q), page);
+}
+
 static ssize_t queue_zone_append_max_show(struct request_queue *q, char *page)
 {
 	unsigned long long max_sectors = q->limits.max_zone_append_sectors;
@@ -585,6 +591,7 @@ QUEUE_RO_ENTRY(queue_discard_zeroes_data, "discard_zeroes_data");
 QUEUE_RO_ENTRY(queue_write_same_max, "write_same_max_bytes");
 QUEUE_RO_ENTRY(queue_write_zeroes_max, "write_zeroes_max_bytes");
 QUEUE_RO_ENTRY(queue_zone_append_max, "zone_append_max_bytes");
+QUEUE_RO_ENTRY(queue_zone_write_granularity, "zone_write_granularity");
 
 QUEUE_RO_ENTRY(queue_zoned, "zoned");
 QUEUE_RO_ENTRY(queue_nr_zones, "nr_zones");
@@ -639,6 +646,7 @@ static struct attribute *queue_attrs[] = {
 	&queue_write_same_max_entry.attr,
 	&queue_write_zeroes_max_entry.attr,
 	&queue_zone_append_max_entry.attr,
+	&queue_zone_write_granularity_entry.attr,
 	&queue_nonrot_entry.attr,
 	&queue_zoned_entry.attr,
 	&queue_nr_zones_entry.attr,
diff --git a/drivers/block/null_blk/zoned.c b/drivers/block/null_blk/zoned.c
index 535351570bb2..704d09481e0d 100644
--- a/drivers/block/null_blk/zoned.c
+++ b/drivers/block/null_blk/zoned.c
@@ -172,6 +172,7 @@ int null_register_zoned_dev(struct nullb *nullb)
 	blk_queue_max_zone_append_sectors(q, dev->zone_size_sects);
 	blk_queue_max_open_zones(q, dev->zone_max_open);
 	blk_queue_max_active_zones(q, dev->zone_max_active);
+	blk_queue_zone_write_granularity(q, dev->blocksize);
 
 	return 0;
 }
diff --git a/drivers/nvme/host/zns.c b/drivers/nvme/host/zns.c
index 1dfe9a3500e3..f25311ccd996 100644
--- a/drivers/nvme/host/zns.c
+++ b/drivers/nvme/host/zns.c
@@ -113,6 +113,7 @@ int nvme_update_zone_info(struct nvme_ns *ns, unsigned lbaf)
 	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);
+	blk_queue_zone_write_granularity(q, q->limits.logical_block_size);
 free_data:
 	kfree(id);
 	return status;
diff --git a/drivers/scsi/sd_zbc.c b/drivers/scsi/sd_zbc.c
index cf07b7f93579..10e9f33cc069 100644
--- a/drivers/scsi/sd_zbc.c
+++ b/drivers/scsi/sd_zbc.c
@@ -789,6 +789,16 @@ int sd_zbc_read_zones(struct scsi_disk *sdkp, unsigned char *buf)
 	blk_queue_max_active_zones(q, 0);
 	nr_zones = round_up(sdkp->capacity, zone_blocks) >> ilog2(zone_blocks);
 
+	/*
+	 * Per ZBC and ZAC specifications, writes in sequential write required
+	 * zones of host-managed devices must be aligned to the device physical
+	 * block size.
+	 */
+	if (blk_queue_zoned_model(q) == BLK_ZONED_HM)
+		blk_queue_zone_write_granularity(q, sdkp->physical_block_size);
+	else if (blk_queue_zoned_model(q) == BLK_ZONED_HA)
+		blk_queue_zone_write_granularity(q, sdkp->device->sector_size);
+
 	/* READ16/WRITE16 is mandatory for ZBC disks */
 	sdkp->device->use_16_for_rw = 1;
 	sdkp->device->use_10_for_rw = 0;
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index f94ee3089e01..142e3b34be75 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -337,6 +337,7 @@ struct queue_limits {
 	unsigned int		max_zone_append_sectors;
 	unsigned int		discard_granularity;
 	unsigned int		discard_alignment;
+	unsigned int		zone_write_granularity;
 
 	unsigned short		max_segments;
 	unsigned short		max_integrity_segments;
@@ -1161,6 +1162,8 @@ extern void blk_queue_logical_block_size(struct request_queue *, unsigned int);
 extern void blk_queue_max_zone_append_sectors(struct request_queue *q,
 		unsigned int max_zone_append_sectors);
 extern void blk_queue_physical_block_size(struct request_queue *, unsigned int);
+void blk_queue_zone_write_granularity(struct request_queue *q,
+				      unsigned int size);
 extern void blk_queue_alignment_offset(struct request_queue *q,
 				       unsigned int alignment);
 void blk_queue_update_readahead(struct request_queue *q);
@@ -1474,6 +1477,18 @@ static inline int bdev_io_opt(struct block_device *bdev)
 	return queue_io_opt(bdev_get_queue(bdev));
 }
 
+static inline unsigned int
+queue_zone_write_granularity(const struct request_queue *q)
+{
+	return q->limits.zone_write_granularity;
+}
+
+static inline unsigned int
+bdev_zone_write_granularity(struct block_device *bdev)
+{
+	return queue_zone_write_granularity(bdev_get_queue(bdev));
+}
+
 static inline int queue_alignment_offset(const struct request_queue *q)
 {
 	if (q->limits.misaligned)
-- 
2.29.2

[PATCH v3 1/3] block: introduce zone_write_granularity limit

[Damien Le Moal]

Per ZBC and ZAC specifications, host-managed SMR hard-disks mandate that
all writes into sequential write required zones be aligned to the device
physical block size. However, NVMe ZNS does not have this constraint and
allows write operations into sequential zones to be logical block size
aligned. This inconsistency does not help with portability of software
across device types.
To solve this, introduce the zone_write_granularity queue limit to
indicate the alignment constraint, in bytes, of write operations into
zones of a zoned block device. This new limit is exported as a
read-only sysfs queue attribute and the helper
blk_queue_zone_write_granularity() introduced for drivers to set this
limit. The scsi disk driver is modified to use this helper to set
host-managed SMR disk zone write granularity to the disk physical block
size. The ZNS support code of the NVMe driver is also modified to use
this helper to set the new limit to the logical block size of the
namespace. The nullblk driver is similarly modified too.

The accessor functions queue_zone_write_granularity() and
bdev_zone_write_granularity() are also introduced.

Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>
---
 Documentation/block/queue-sysfs.rst |  7 ++++++
 block/blk-settings.c                | 39 +++++++++++++++++++++++++----
 block/blk-sysfs.c                   |  8 ++++++
 drivers/block/null_blk/zoned.c      |  1 +
 drivers/nvme/host/zns.c             |  1 +
 drivers/scsi/sd_zbc.c               | 10 ++++++++
 include/linux/blkdev.h              | 15 +++++++++++
 7 files changed, 76 insertions(+), 5 deletions(-)

diff --git a/Documentation/block/queue-sysfs.rst b/Documentation/block/queue-sysfs.rst
index 2638d3446b79..c8bf8bc3c03a 100644
--- a/Documentation/block/queue-sysfs.rst
+++ b/Documentation/block/queue-sysfs.rst
@@ -273,4 +273,11 @@ devices are described in the ZBC (Zoned Block Commands) and ZAC
 do not support zone commands, they will be treated as regular block devices
 and zoned will report "none".
 
+zone_write_granularity (RO)
+---------------------------
+This indicates the alignment constraint, in bytes, for write operations in
+sequential zones of zoned block devices (devices with a zoned attributed
+that reports "host-managed" or "host-aware"). This value is always 0 for
+regular block devices.
+
 Jens Axboe <jens.axboe@oracle.com>, February 2009
diff --git a/block/blk-settings.c b/block/blk-settings.c
index 43990b1d148b..48872b4085d4 100644
--- a/block/blk-settings.c
+++ b/block/blk-settings.c
@@ -60,6 +60,7 @@ void blk_set_default_limits(struct queue_limits *lim)
 	lim->io_opt = 0;
 	lim->misaligned = 0;
 	lim->zoned = BLK_ZONED_NONE;
+	lim->zone_write_granularity = 0;
 }
 EXPORT_SYMBOL(blk_set_default_limits);
 
@@ -366,6 +367,28 @@ void blk_queue_physical_block_size(struct request_queue *q, unsigned int size)
 }
 EXPORT_SYMBOL(blk_queue_physical_block_size);
 
+/**
+ * blk_queue_zone_write_granularity - set zone write granularity for the queue
+ * @q:  the request queue for the zoned device
+ * @size:  the zone write granularity size, in bytes
+ *
+ * Description:
+ *   This should be set to the lowest possible size allowing to write in
+ *   sequential zones of a zoned block device.
+ */
+void blk_queue_zone_write_granularity(struct request_queue *q,
+				      unsigned int size)
+{
+	if (WARN_ON_ONCE(!blk_queue_is_zoned(q)))
+		return;
+
+	q->limits.zone_write_granularity = size;
+
+	if (q->limits.zone_write_granularity < q->limits.logical_block_size)
+		q->limits.zone_write_granularity = q->limits.logical_block_size;
+}
+EXPORT_SYMBOL_GPL(blk_queue_zone_write_granularity);
+
 /**
  * blk_queue_alignment_offset - set physical block alignment offset
  * @q:	the request queue for the device
@@ -631,6 +654,8 @@ int blk_stack_limits(struct queue_limits *t, struct queue_limits *b,
 			t->discard_granularity;
 	}
 
+	t->zone_write_granularity = max(t->zone_write_granularity,
+					b->zone_write_granularity);
 	t->zoned = max(t->zoned, b->zoned);
 	return ret;
 }
@@ -847,6 +872,8 @@ EXPORT_SYMBOL_GPL(blk_queue_can_use_dma_map_merging);
  */
 void blk_queue_set_zoned(struct gendisk *disk, enum blk_zoned_model model)
 {
+	struct request_queue *q = disk->queue;
+
 	switch (model) {
 	case BLK_ZONED_HM:
 		/*
@@ -864,18 +891,20 @@ void blk_queue_set_zoned(struct gendisk *disk, enum blk_zoned_model model)
 		 * partitions and zoned block device support is enabled, else
 		 * we do nothing special as far as the block layer is concerned.
 		 */
-		if (!IS_ENABLED(CONFIG_BLK_DEV_ZONED) ||
-		    disk_has_partitions(disk))
-			model = BLK_ZONED_NONE;
-		break;
+		if (IS_ENABLED(CONFIG_BLK_DEV_ZONED) &&
+		    !disk_has_partitions(disk))
+			break;
+		model = BLK_ZONED_NONE;
+		fallthrough;
 	case BLK_ZONED_NONE:
 	default:
 		if (WARN_ON_ONCE(model != BLK_ZONED_NONE))
 			model = BLK_ZONED_NONE;
+		q->limits.zone_write_granularity = 0;
 		break;
 	}
 
-	disk->queue->limits.zoned = model;
+	q->limits.zoned = model;
 }
 EXPORT_SYMBOL_GPL(blk_queue_set_zoned);
 
diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c
index b513f1683af0..ae39c7f3d83d 100644
--- a/block/blk-sysfs.c
+++ b/block/blk-sysfs.c
@@ -219,6 +219,12 @@ static ssize_t queue_write_zeroes_max_show(struct request_queue *q, char *page)
 		(unsigned long long)q->limits.max_write_zeroes_sectors << 9);
 }
 
+static ssize_t queue_zone_write_granularity_show(struct request_queue *q,
+						 char *page)
+{
+	return queue_var_show(queue_zone_write_granularity(q), page);
+}
+
 static ssize_t queue_zone_append_max_show(struct request_queue *q, char *page)
 {
 	unsigned long long max_sectors = q->limits.max_zone_append_sectors;
@@ -585,6 +591,7 @@ QUEUE_RO_ENTRY(queue_discard_zeroes_data, "discard_zeroes_data");
 QUEUE_RO_ENTRY(queue_write_same_max, "write_same_max_bytes");
 QUEUE_RO_ENTRY(queue_write_zeroes_max, "write_zeroes_max_bytes");
 QUEUE_RO_ENTRY(queue_zone_append_max, "zone_append_max_bytes");
+QUEUE_RO_ENTRY(queue_zone_write_granularity, "zone_write_granularity");
 
 QUEUE_RO_ENTRY(queue_zoned, "zoned");
 QUEUE_RO_ENTRY(queue_nr_zones, "nr_zones");
@@ -639,6 +646,7 @@ static struct attribute *queue_attrs[] = {
 	&queue_write_same_max_entry.attr,
 	&queue_write_zeroes_max_entry.attr,
 	&queue_zone_append_max_entry.attr,
+	&queue_zone_write_granularity_entry.attr,
 	&queue_nonrot_entry.attr,
 	&queue_zoned_entry.attr,
 	&queue_nr_zones_entry.attr,
diff --git a/drivers/block/null_blk/zoned.c b/drivers/block/null_blk/zoned.c
index 535351570bb2..704d09481e0d 100644
--- a/drivers/block/null_blk/zoned.c
+++ b/drivers/block/null_blk/zoned.c
@@ -172,6 +172,7 @@ int null_register_zoned_dev(struct nullb *nullb)
 	blk_queue_max_zone_append_sectors(q, dev->zone_size_sects);
 	blk_queue_max_open_zones(q, dev->zone_max_open);
 	blk_queue_max_active_zones(q, dev->zone_max_active);
+	blk_queue_zone_write_granularity(q, dev->blocksize);
 
 	return 0;
 }
diff --git a/drivers/nvme/host/zns.c b/drivers/nvme/host/zns.c
index 1dfe9a3500e3..f25311ccd996 100644
--- a/drivers/nvme/host/zns.c
+++ b/drivers/nvme/host/zns.c
@@ -113,6 +113,7 @@ int nvme_update_zone_info(struct nvme_ns *ns, unsigned lbaf)
 	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);
+	blk_queue_zone_write_granularity(q, q->limits.logical_block_size);
 free_data:
 	kfree(id);
 	return status;
diff --git a/drivers/scsi/sd_zbc.c b/drivers/scsi/sd_zbc.c
index cf07b7f93579..10e9f33cc069 100644
--- a/drivers/scsi/sd_zbc.c
+++ b/drivers/scsi/sd_zbc.c
@@ -789,6 +789,16 @@ int sd_zbc_read_zones(struct scsi_disk *sdkp, unsigned char *buf)
 	blk_queue_max_active_zones(q, 0);
 	nr_zones = round_up(sdkp->capacity, zone_blocks) >> ilog2(zone_blocks);
 
+	/*
+	 * Per ZBC and ZAC specifications, writes in sequential write required
+	 * zones of host-managed devices must be aligned to the device physical
+	 * block size.
+	 */
+	if (blk_queue_zoned_model(q) == BLK_ZONED_HM)
+		blk_queue_zone_write_granularity(q, sdkp->physical_block_size);
+	else if (blk_queue_zoned_model(q) == BLK_ZONED_HA)
+		blk_queue_zone_write_granularity(q, sdkp->device->sector_size);
+
 	/* READ16/WRITE16 is mandatory for ZBC disks */
 	sdkp->device->use_16_for_rw = 1;
 	sdkp->device->use_10_for_rw = 0;
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index f94ee3089e01..142e3b34be75 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -337,6 +337,7 @@ struct queue_limits {
 	unsigned int		max_zone_append_sectors;
 	unsigned int		discard_granularity;
 	unsigned int		discard_alignment;
+	unsigned int		zone_write_granularity;
 
 	unsigned short		max_segments;
 	unsigned short		max_integrity_segments;
@@ -1161,6 +1162,8 @@ extern void blk_queue_logical_block_size(struct request_queue *, unsigned int);
 extern void blk_queue_max_zone_append_sectors(struct request_queue *q,
 		unsigned int max_zone_append_sectors);
 extern void blk_queue_physical_block_size(struct request_queue *, unsigned int);
+void blk_queue_zone_write_granularity(struct request_queue *q,
+				      unsigned int size);
 extern void blk_queue_alignment_offset(struct request_queue *q,
 				       unsigned int alignment);
 void blk_queue_update_readahead(struct request_queue *q);
@@ -1474,6 +1477,18 @@ static inline int bdev_io_opt(struct block_device *bdev)
 	return queue_io_opt(bdev_get_queue(bdev));
 }
 
+static inline unsigned int
+queue_zone_write_granularity(const struct request_queue *q)
+{
+	return q->limits.zone_write_granularity;
+}
+
+static inline unsigned int
+bdev_zone_write_granularity(struct block_device *bdev)
+{
+	return queue_zone_write_granularity(bdev_get_queue(bdev));
+}
+
 static inline int queue_alignment_offset(const struct request_queue *q)
 {
 	if (q->limits.misaligned)
-- 
2.29.2


_______________________________________________
Linux-nvme mailing list
Linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme

[PATCH v3 2/3] block: document zone_append_max_bytes attribute

[Damien Le Moal]

The description of the zone_append_max_bytes sysfs queue attribute is
missing from Documentation/block/queue-sysfs.rst. Add it.

Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 Documentation/block/queue-sysfs.rst | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/Documentation/block/queue-sysfs.rst b/Documentation/block/queue-sysfs.rst
index c8bf8bc3c03a..4dc7f0d499a8 100644
--- a/Documentation/block/queue-sysfs.rst
+++ b/Documentation/block/queue-sysfs.rst
@@ -261,6 +261,12 @@ For block drivers that support REQ_OP_WRITE_ZEROES, the maximum number of
 bytes that can be zeroed at once. The value 0 means that REQ_OP_WRITE_ZEROES
 is not supported.
 
+zone_append_max_bytes (RO)
+--------------------------
+This is the maximum number of bytes that can be written to a sequential
+zone of a zoned block device using a zone append write operation
+(REQ_OP_ZONE_APPEND). This value is always 0 for regular block devices.
+
 zoned (RO)
 ----------
 This indicates if the device is a zoned block device and the zone model of the
-- 
2.29.2

[PATCH v3 2/3] block: document zone_append_max_bytes attribute

[Damien Le Moal]

The description of the zone_append_max_bytes sysfs queue attribute is
missing from Documentation/block/queue-sysfs.rst. Add it.

Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 Documentation/block/queue-sysfs.rst | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/Documentation/block/queue-sysfs.rst b/Documentation/block/queue-sysfs.rst
index c8bf8bc3c03a..4dc7f0d499a8 100644
--- a/Documentation/block/queue-sysfs.rst
+++ b/Documentation/block/queue-sysfs.rst
@@ -261,6 +261,12 @@ For block drivers that support REQ_OP_WRITE_ZEROES, the maximum number of
 bytes that can be zeroed at once. The value 0 means that REQ_OP_WRITE_ZEROES
 is not supported.
 
+zone_append_max_bytes (RO)
+--------------------------
+This is the maximum number of bytes that can be written to a sequential
+zone of a zoned block device using a zone append write operation
+(REQ_OP_ZONE_APPEND). This value is always 0 for regular block devices.
+
 zoned (RO)
 ----------
 This indicates if the device is a zoned block device and the zone model of the
-- 
2.29.2


_______________________________________________
Linux-nvme mailing list
Linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme

[PATCH v3 3/3] zonefs: use zone write granularity as block size

[Damien Le Moal]

Zoned block devices have different granularity constraints for write
operations into sequential zones. E.g. ZBC and ZAC devices require that
writes be aligned to the device physical block size while NVMe ZNS
devices allow logical block size aligned write operations. To correctly
handle such difference, use the device zone write granularity limit to
set the block size of a zonefs volume, thus allowing the smallest
possible write unit for all zoned device types.

Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>
---
 fs/zonefs/super.c | 9 ++++-----
 1 file changed, 4 insertions(+), 5 deletions(-)

diff --git a/fs/zonefs/super.c b/fs/zonefs/super.c
index bec47f2d074b..8973d77ba000 100644
--- a/fs/zonefs/super.c
+++ b/fs/zonefs/super.c
@@ -1581,12 +1581,11 @@ static int zonefs_fill_super(struct super_block *sb, void *data, int silent)
 	sb->s_time_gran	= 1;
 
 	/*
-	 * The block size is set to the device physical sector size to ensure
-	 * that write operations on 512e devices (512B logical block and 4KB
-	 * physical block) are always aligned to the device physical blocks,
-	 * as mandated by the ZBC/ZAC specifications.
+	 * The block size is set to the device zone write granularity to ensure
+	 * that write operations are always aligned according to the device
+	 * interface constraints.
 	 */
-	sb_set_blocksize(sb, bdev_physical_block_size(sb->s_bdev));
+	sb_set_blocksize(sb, bdev_zone_write_granularity(sb->s_bdev));
 	sbi->s_zone_sectors_shift = ilog2(bdev_zone_sectors(sb->s_bdev));
 	sbi->s_uid = GLOBAL_ROOT_UID;
 	sbi->s_gid = GLOBAL_ROOT_GID;
-- 
2.29.2

[PATCH v3 3/3] zonefs: use zone write granularity as block size

[Damien Le Moal]

Zoned block devices have different granularity constraints for write
operations into sequential zones. E.g. ZBC and ZAC devices require that
writes be aligned to the device physical block size while NVMe ZNS
devices allow logical block size aligned write operations. To correctly
handle such difference, use the device zone write granularity limit to
set the block size of a zonefs volume, thus allowing the smallest
possible write unit for all zoned device types.

Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>
---
 fs/zonefs/super.c | 9 ++++-----
 1 file changed, 4 insertions(+), 5 deletions(-)

diff --git a/fs/zonefs/super.c b/fs/zonefs/super.c
index bec47f2d074b..8973d77ba000 100644
--- a/fs/zonefs/super.c
+++ b/fs/zonefs/super.c
@@ -1581,12 +1581,11 @@ static int zonefs_fill_super(struct super_block *sb, void *data, int silent)
 	sb->s_time_gran	= 1;
 
 	/*
-	 * The block size is set to the device physical sector size to ensure
-	 * that write operations on 512e devices (512B logical block and 4KB
-	 * physical block) are always aligned to the device physical blocks,
-	 * as mandated by the ZBC/ZAC specifications.
+	 * The block size is set to the device zone write granularity to ensure
+	 * that write operations are always aligned according to the device
+	 * interface constraints.
 	 */
-	sb_set_blocksize(sb, bdev_physical_block_size(sb->s_bdev));
+	sb_set_blocksize(sb, bdev_zone_write_granularity(sb->s_bdev));
 	sbi->s_zone_sectors_shift = ilog2(bdev_zone_sectors(sb->s_bdev));
 	sbi->s_uid = GLOBAL_ROOT_UID;
 	sbi->s_gid = GLOBAL_ROOT_GID;
-- 
2.29.2


_______________________________________________
Linux-nvme mailing list
Linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme

Re: [PATCH v3 1/3] block: introduce zone_write_granularity limit

[Christoph Hellwig]

> @@ -864,18 +891,20 @@ void blk_queue_set_zoned(struct gendisk *disk, enum blk_zoned_model model)
>  		 * partitions and zoned block device support is enabled, else
>  		 * we do nothing special as far as the block layer is concerned.
>  		 */
> -		if (!IS_ENABLED(CONFIG_BLK_DEV_ZONED) ||
> -		    disk_has_partitions(disk))
> -			model = BLK_ZONED_NONE;
> -		break;
> +		if (IS_ENABLED(CONFIG_BLK_DEV_ZONED) &&
> +		    !disk_has_partitions(disk))
> +			break;
> +		model = BLK_ZONED_NONE;
> +		fallthrough;
>  	case BLK_ZONED_NONE:
>  	default:
>  		if (WARN_ON_ONCE(model != BLK_ZONED_NONE))
>  			model = BLK_ZONED_NONE;
> +		q->limits.zone_write_granularity = 0;
>  		break;
>  	}
>  
> -	disk->queue->limits.zoned = model;
> +	q->limits.zoned = model;
>  }

This looks a little strange.  If we special case zoned vs not zoned
here anyway, why not set the zone_write_granularity to the logical
block size here by default.

Re: [PATCH v3 1/3] block: introduce zone_write_granularity limit

[Christoph Hellwig]

> @@ -864,18 +891,20 @@ void blk_queue_set_zoned(struct gendisk *disk, enum blk_zoned_model model)
>  		 * partitions and zoned block device support is enabled, else
>  		 * we do nothing special as far as the block layer is concerned.
>  		 */
> -		if (!IS_ENABLED(CONFIG_BLK_DEV_ZONED) ||
> -		    disk_has_partitions(disk))
> -			model = BLK_ZONED_NONE;
> -		break;
> +		if (IS_ENABLED(CONFIG_BLK_DEV_ZONED) &&
> +		    !disk_has_partitions(disk))
> +			break;
> +		model = BLK_ZONED_NONE;
> +		fallthrough;
>  	case BLK_ZONED_NONE:
>  	default:
>  		if (WARN_ON_ONCE(model != BLK_ZONED_NONE))
>  			model = BLK_ZONED_NONE;
> +		q->limits.zone_write_granularity = 0;
>  		break;
>  	}
>  
> -	disk->queue->limits.zoned = model;
> +	q->limits.zoned = model;
>  }

This looks a little strange.  If we special case zoned vs not zoned
here anyway, why not set the zone_write_granularity to the logical
block size here by default.

_______________________________________________
Linux-nvme mailing list
Linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme

Re: [PATCH v3 3/3] zonefs: use zone write granularity as block size

[Christoph Hellwig]

Looks good,

Reviewed-by: Christoph Hellwig <hch@lst.de>

Re: [PATCH v3 3/3] zonefs: use zone write granularity as block size

[Christoph Hellwig]

Looks good,

Reviewed-by: Christoph Hellwig <hch@lst.de>

_______________________________________________
Linux-nvme mailing list
Linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme

Re: [PATCH v3 1/3] block: introduce zone_write_granularity limit

[Damien Le Moal]

On 2021/01/22 17:42, Christoph Hellwig wrote:
>> @@ -864,18 +891,20 @@ void blk_queue_set_zoned(struct gendisk *disk, enum blk_zoned_model model)
>>  		 * partitions and zoned block device support is enabled, else
>>  		 * we do nothing special as far as the block layer is concerned.
>>  		 */
>> -		if (!IS_ENABLED(CONFIG_BLK_DEV_ZONED) ||
>> -		    disk_has_partitions(disk))
>> -			model = BLK_ZONED_NONE;
>> -		break;
>> +		if (IS_ENABLED(CONFIG_BLK_DEV_ZONED) &&
>> +		    !disk_has_partitions(disk))
>> +			break;
>> +		model = BLK_ZONED_NONE;
>> +		fallthrough;
>>  	case BLK_ZONED_NONE:
>>  	default:
>>  		if (WARN_ON_ONCE(model != BLK_ZONED_NONE))
>>  			model = BLK_ZONED_NONE;
>> +		q->limits.zone_write_granularity = 0;
>>  		break;
>>  	}
>>  
>> -	disk->queue->limits.zoned = model;
>> +	q->limits.zoned = model;
>>  }
> 
> This looks a little strange.  If we special case zoned vs not zoned
> here anyway, why not set the zone_write_granularity to the logical
> block size here by default.

The convention is zone_write_granularity == 0 for the BLK_ZONED_NONE case. Hence
the reset here if we force the zoned model to none for HA drives. This way, this
does not create a special case for HA drives used as regular disks.

Of note is that there is something a little weird in the sd_zbc.c code that
needs fixing: blk_queue_set_zoned() is called before sd_zbc_read_zones() is
executed and that function will check the zones of an HA drive and set the queue
nr_zones and max zone append sectors, even if blk_queue_set_zoned() set the
zoned model to none due to partitions. That makes the BLK_ZONED_NONE case of HA
drives a little weird since zone information is visible and correct but the
model says "none". As long as users separate zoned vs not-zoned cases by looking
at the zoned model, this does not create any problem, but that is not pretty.

Will send a separate patch to clean that up and have something consistent with
regular disks for this special HA case. The above blk_queue_set_zoned() function
can be used to cleanup the zones information for an HA drive that is used as a
regular disk (nr_zones, zone append sectors and zone bitmaps).


-- 
Damien Le Moal
Western Digital Research

Re: [PATCH v3 1/3] block: introduce zone_write_granularity limit

[Damien Le Moal]

On 2021/01/22 17:42, Christoph Hellwig wrote:
>> @@ -864,18 +891,20 @@ void blk_queue_set_zoned(struct gendisk *disk, enum blk_zoned_model model)
>>  		 * partitions and zoned block device support is enabled, else
>>  		 * we do nothing special as far as the block layer is concerned.
>>  		 */
>> -		if (!IS_ENABLED(CONFIG_BLK_DEV_ZONED) ||
>> -		    disk_has_partitions(disk))
>> -			model = BLK_ZONED_NONE;
>> -		break;
>> +		if (IS_ENABLED(CONFIG_BLK_DEV_ZONED) &&
>> +		    !disk_has_partitions(disk))
>> +			break;
>> +		model = BLK_ZONED_NONE;
>> +		fallthrough;
>>  	case BLK_ZONED_NONE:
>>  	default:
>>  		if (WARN_ON_ONCE(model != BLK_ZONED_NONE))
>>  			model = BLK_ZONED_NONE;
>> +		q->limits.zone_write_granularity = 0;
>>  		break;
>>  	}
>>  
>> -	disk->queue->limits.zoned = model;
>> +	q->limits.zoned = model;
>>  }
> 
> This looks a little strange.  If we special case zoned vs not zoned
> here anyway, why not set the zone_write_granularity to the logical
> block size here by default.

The convention is zone_write_granularity == 0 for the BLK_ZONED_NONE case. Hence
the reset here if we force the zoned model to none for HA drives. This way, this
does not create a special case for HA drives used as regular disks.

Of note is that there is something a little weird in the sd_zbc.c code that
needs fixing: blk_queue_set_zoned() is called before sd_zbc_read_zones() is
executed and that function will check the zones of an HA drive and set the queue
nr_zones and max zone append sectors, even if blk_queue_set_zoned() set the
zoned model to none due to partitions. That makes the BLK_ZONED_NONE case of HA
drives a little weird since zone information is visible and correct but the
model says "none". As long as users separate zoned vs not-zoned cases by looking
at the zoned model, this does not create any problem, but that is not pretty.

Will send a separate patch to clean that up and have something consistent with
regular disks for this special HA case. The above blk_queue_set_zoned() function
can be used to cleanup the zones information for an HA drive that is used as a
regular disk (nr_zones, zone append sectors and zone bitmaps).


-- 
Damien Le Moal
Western Digital Research

_______________________________________________
Linux-nvme mailing list
Linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme

Re: [PATCH v3 1/3] block: introduce zone_write_granularity limit

[Martin K. Petersen]

Damien,

> To solve this, introduce the zone_write_granularity queue limit to
> indicate the alignment constraint, in bytes, of write operations into
> zones of a zoned block device. This new limit is exported as a
> read-only sysfs queue attribute and the helper
> blk_queue_zone_write_granularity() introduced for drivers to set this
> limit. The scsi disk driver is modified to use this helper to set
> host-managed SMR disk zone write granularity to the disk physical
> block size. The ZNS support code of the NVMe driver is also modified
> to use this helper to set the new limit to the logical block size of
> the namespace. The nullblk driver is similarly modified too.

Looks fine.

Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>

-- 
Martin K. Petersen	Oracle Linux Engineering

Re: [PATCH v3 1/3] block: introduce zone_write_granularity limit

[Martin K. Petersen]

Damien,

> To solve this, introduce the zone_write_granularity queue limit to
> indicate the alignment constraint, in bytes, of write operations into
> zones of a zoned block device. This new limit is exported as a
> read-only sysfs queue attribute and the helper
> blk_queue_zone_write_granularity() introduced for drivers to set this
> limit. The scsi disk driver is modified to use this helper to set
> host-managed SMR disk zone write granularity to the disk physical
> block size. The ZNS support code of the NVMe driver is also modified
> to use this helper to set the new limit to the logical block size of
> the namespace. The nullblk driver is similarly modified too.

Looks fine.

Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>

-- 
Martin K. Petersen	Oracle Linux Engineering

_______________________________________________
Linux-nvme mailing list
Linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme

Re: [PATCH v3 2/3] block: document zone_append_max_bytes attribute

[Martin K. Petersen]

Damien,

> The description of the zone_append_max_bytes sysfs queue attribute is
> missing from Documentation/block/queue-sysfs.rst. Add it.

Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>

-- 
Martin K. Petersen	Oracle Linux Engineering

Re: [PATCH v3 2/3] block: document zone_append_max_bytes attribute

[Martin K. Petersen]

Damien,

> The description of the zone_append_max_bytes sysfs queue attribute is
> missing from Documentation/block/queue-sysfs.rst. Add it.

Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>

-- 
Martin K. Petersen	Oracle Linux Engineering

_______________________________________________
Linux-nvme mailing list
Linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme

Re: [PATCH v3 3/3] zonefs: use zone write granularity as block size

[Martin K. Petersen]

Damien,

> Zoned block devices have different granularity constraints for write
> operations into sequential zones. E.g. ZBC and ZAC devices require
> that writes be aligned to the device physical block size while NVMe
> ZNS devices allow logical block size aligned write operations. To
> correctly handle such difference, use the device zone write
> granularity limit to set the block size of a zonefs volume, thus
> allowing the smallest possible write unit for all zoned device types.

Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>

-- 
Martin K. Petersen	Oracle Linux Engineering

Re: [PATCH v3 3/3] zonefs: use zone write granularity as block size

[Martin K. Petersen]

Damien,

> Zoned block devices have different granularity constraints for write
> operations into sequential zones. E.g. ZBC and ZAC devices require
> that writes be aligned to the device physical block size while NVMe
> ZNS devices allow logical block size aligned write operations. To
> correctly handle such difference, use the device zone write
> granularity limit to set the block size of a zonefs volume, thus
> allowing the smallest possible write unit for all zoned device types.

Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>

-- 
Martin K. Petersen	Oracle Linux Engineering

_______________________________________________
Linux-nvme mailing list
Linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme

Re: [PATCH v3 2/3] block: document zone_append_max_bytes attribute

[Chaitanya Kulkarni]

On 1/22/21 12:00 AM, Damien Le Moal wrote:
> The description of the zone_append_max_bytes sysfs queue attribute is
> missing from Documentation/block/queue-sysfs.rst. Add it.
>
> Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>
> Reviewed-by: Christoph Hellwig <hch@lst.de>
Looks good.

Reviewed-by: Chaitanya Kulkarni <chaitanya.kulkarni@wdc.com>

Re: [PATCH v3 2/3] block: document zone_append_max_bytes attribute

[Chaitanya Kulkarni]

On 1/22/21 12:00 AM, Damien Le Moal wrote:
> The description of the zone_append_max_bytes sysfs queue attribute is
> missing from Documentation/block/queue-sysfs.rst. Add it.
>
> Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>
> Reviewed-by: Christoph Hellwig <hch@lst.de>
Looks good.

Reviewed-by: Chaitanya Kulkarni <chaitanya.kulkarni@wdc.com>

_______________________________________________
Linux-nvme mailing list
Linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme

Re: [PATCH v3 1/3] block: introduce zone_write_granularity limit

[Christoph Hellwig]

On Fri, Jan 22, 2021 at 08:56:58AM +0000, Damien Le Moal wrote:
> > This looks a little strange.  If we special case zoned vs not zoned
> > here anyway, why not set the zone_write_granularity to the logical
> > block size here by default.
> 
> The convention is zone_write_granularity == 0 for the BLK_ZONED_NONE case. Hence
> the reset here if we force the zoned model to none for HA drives. This way, this
> does not create a special case for HA drives used as regular disks.

Just inititialize it for all cases if you initialize it for some here.
That way everyone but sd already gets a right default and life becomes
simpler.

Re: [PATCH v3 1/3] block: introduce zone_write_granularity limit

[Christoph Hellwig]

On Fri, Jan 22, 2021 at 08:56:58AM +0000, Damien Le Moal wrote:
> > This looks a little strange.  If we special case zoned vs not zoned
> > here anyway, why not set the zone_write_granularity to the logical
> > block size here by default.
> 
> The convention is zone_write_granularity == 0 for the BLK_ZONED_NONE case. Hence
> the reset here if we force the zoned model to none for HA drives. This way, this
> does not create a special case for HA drives used as regular disks.

Just inititialize it for all cases if you initialize it for some here.
That way everyone but sd already gets a right default and life becomes
simpler.

_______________________________________________
Linux-nvme mailing list
Linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme

Re: [PATCH v3 1/3] block: introduce zone_write_granularity limit

[Damien Le Moal]

On 2021/01/24 19:07, Christoph Hellwig wrote:
> On Fri, Jan 22, 2021 at 08:56:58AM +0000, Damien Le Moal wrote:
>>> This looks a little strange.  If we special case zoned vs not zoned
>>> here anyway, why not set the zone_write_granularity to the logical
>>> block size here by default.
>>
>> The convention is zone_write_granularity == 0 for the BLK_ZONED_NONE case. Hence
>> the reset here if we force the zoned model to none for HA drives. This way, this
>> does not create a special case for HA drives used as regular disks.
> 
> Just inititialize it for all cases if you initialize it for some here.
> That way everyone but sd already gets a right default and life becomes
> simpler.

True for nullblk, and that also simplifies sd a little. But not for nvme,
blk_queue_set_zoned() is not used AND nvme_update_zone_info() is called before
nvme_update_disk_info() where the NS logical block size is set. So some
surgery/cleanups would be needed to benefit. I could add a cleanup for this, but
not entirely sure if calling nvme_update_zone_info() after
nvme_update_disk_info() is OK. Thoughts ?


-- 
Damien Le Moal
Western Digital Research

Re: [PATCH v3 1/3] block: introduce zone_write_granularity limit

[Damien Le Moal]

On 2021/01/24 19:07, Christoph Hellwig wrote:
> On Fri, Jan 22, 2021 at 08:56:58AM +0000, Damien Le Moal wrote:
>>> This looks a little strange.  If we special case zoned vs not zoned
>>> here anyway, why not set the zone_write_granularity to the logical
>>> block size here by default.
>>
>> The convention is zone_write_granularity == 0 for the BLK_ZONED_NONE case. Hence
>> the reset here if we force the zoned model to none for HA drives. This way, this
>> does not create a special case for HA drives used as regular disks.
> 
> Just inititialize it for all cases if you initialize it for some here.
> That way everyone but sd already gets a right default and life becomes
> simpler.

True for nullblk, and that also simplifies sd a little. But not for nvme,
blk_queue_set_zoned() is not used AND nvme_update_zone_info() is called before
nvme_update_disk_info() where the NS logical block size is set. So some
surgery/cleanups would be needed to benefit. I could add a cleanup for this, but
not entirely sure if calling nvme_update_zone_info() after
nvme_update_disk_info() is OK. Thoughts ?


-- 
Damien Le Moal
Western Digital Research

_______________________________________________
Linux-nvme mailing list
Linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme

Re: [PATCH v3 1/3] block: introduce zone_write_granularity limit

[Damien Le Moal]

On 2021/01/25 14:32, Damien Le Moal wrote:
> On 2021/01/24 19:07, Christoph Hellwig wrote:
>> On Fri, Jan 22, 2021 at 08:56:58AM +0000, Damien Le Moal wrote:
>>>> This looks a little strange.  If we special case zoned vs not zoned
>>>> here anyway, why not set the zone_write_granularity to the logical
>>>> block size here by default.
>>>
>>> The convention is zone_write_granularity == 0 for the BLK_ZONED_NONE case. Hence
>>> the reset here if we force the zoned model to none for HA drives. This way, this
>>> does not create a special case for HA drives used as regular disks.
>>
>> Just inititialize it for all cases if you initialize it for some here.
>> That way everyone but sd already gets a right default and life becomes
>> simpler.
> 
> True for nullblk, and that also simplifies sd a little. But not for nvme,
> blk_queue_set_zoned() is not used AND nvme_update_zone_info() is called before
> nvme_update_disk_info() where the NS logical block size is set. So some
> surgery/cleanups would be needed to benefit. I could add a cleanup for this, but
> not entirely sure if calling nvme_update_zone_info() after
> nvme_update_disk_info() is OK. Thoughts ?

Also, because nvme_update_zone_info() is called without the logical block size
being set yet, this patch was not good for ZNS.




-- 
Damien Le Moal
Western Digital Research

Re: [PATCH v3 1/3] block: introduce zone_write_granularity limit

[Damien Le Moal]

On 2021/01/25 14:32, Damien Le Moal wrote:
> On 2021/01/24 19:07, Christoph Hellwig wrote:
>> On Fri, Jan 22, 2021 at 08:56:58AM +0000, Damien Le Moal wrote:
>>>> This looks a little strange.  If we special case zoned vs not zoned
>>>> here anyway, why not set the zone_write_granularity to the logical
>>>> block size here by default.
>>>
>>> The convention is zone_write_granularity == 0 for the BLK_ZONED_NONE case. Hence
>>> the reset here if we force the zoned model to none for HA drives. This way, this
>>> does not create a special case for HA drives used as regular disks.
>>
>> Just inititialize it for all cases if you initialize it for some here.
>> That way everyone but sd already gets a right default and life becomes
>> simpler.
> 
> True for nullblk, and that also simplifies sd a little. But not for nvme,
> blk_queue_set_zoned() is not used AND nvme_update_zone_info() is called before
> nvme_update_disk_info() where the NS logical block size is set. So some
> surgery/cleanups would be needed to benefit. I could add a cleanup for this, but
> not entirely sure if calling nvme_update_zone_info() after
> nvme_update_disk_info() is OK. Thoughts ?

Also, because nvme_update_zone_info() is called without the logical block size
being set yet, this patch was not good for ZNS.




-- 
Damien Le Moal
Western Digital Research

_______________________________________________
Linux-nvme mailing list
Linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme