diff mbox series

[v11,08/40] btrfs: emulated zoned mode on non-zoned devices

Message ID e2bcb873196a16b05d5757cd8087900d4f464347.1608608848.git.naohiro.aota@wdc.com (mailing list archive)
State New, archived
Headers show
Series btrfs: zoned block device support | expand

Commit Message

Naohiro Aota Dec. 22, 2020, 3:49 a.m. UTC
From: Johannes Thumshirn <johannes.thumshirn@wdc.com>

Emulate zoned btrfs mode on non-zoned devices. This is done by "slicing
up" the block-device into static sized chunks and fake a conventional zone
on each of them. The emulated zone size is determined from the size of
device extent.

This is mainly aimed at testing parts of the zoned mode, i.e. the zoned
chunk allocator, on regular block devices.

Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
---
 fs/btrfs/dev-replace.c |   3 +
 fs/btrfs/volumes.c     |  14 +++++
 fs/btrfs/volumes.h     |   3 +
 fs/btrfs/zoned.c       | 121 +++++++++++++++++++++++++++++++++++++----
 fs/btrfs/zoned.h       |  14 +++--
 5 files changed, 139 insertions(+), 16 deletions(-)

Comments

Josef Bacik Jan. 11, 2021, 8:54 p.m. UTC | #1
On 12/21/20 10:49 PM, Naohiro Aota wrote:
> From: Johannes Thumshirn <johannes.thumshirn@wdc.com>
> 
> Emulate zoned btrfs mode on non-zoned devices. This is done by "slicing
> up" the block-device into static sized chunks and fake a conventional zone
> on each of them. The emulated zone size is determined from the size of
> device extent.
> 
> This is mainly aimed at testing parts of the zoned mode, i.e. the zoned
> chunk allocator, on regular block devices.
> 
> Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
> Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>

Reviewed-by: Josef Bacik <josef@toxicpanda.com>

Thanks,

Josef
David Sterba Jan. 13, 2021, 5:58 p.m. UTC | #2
On Tue, Dec 22, 2020 at 12:49:01PM +0900, Naohiro Aota wrote:
> @@ -296,12 +383,22 @@ int btrfs_get_dev_zone_info(struct btrfs_device *device)
>  
>  	device->zone_info = zone_info;
>  
> -	/* device->fs_info is not safe to use for printing messages */
> -	btrfs_info_in_rcu(NULL,

NULL instead of fs_info

> -			"host-%s zoned block device %s, %u zones of %llu bytes",
> -			bdev_zoned_model(bdev) == BLK_ZONED_HM ? "managed" : "aware",
> -			rcu_str_deref(device->name), zone_info->nr_zones,
> -			zone_info->zone_size);
> +	if (bdev_zoned_model(bdev) == BLK_ZONED_HM) {
> +		model = "host-managed zoned";
> +		emulated = "";
> +	} else if (bdev_zoned_model(bdev) == BLK_ZONED_HA) {
> +		model = "host-aware zoned";
> +		emulated = "";
> +	} else if (bdev_zoned_model(bdev) == BLK_ZONED_NONE &&
> +		 device->force_zoned) {
> +		model = "regular";
> +		emulated = "emulated ";
> +	}
> +
> +	btrfs_info_in_rcu(device->fs_info,

so what changed that it's fine to use device->fs_info now while it was
not before?

> +		"%s block device %s, %u %szones of %llu bytes",
> +		model, rcu_str_deref(device->name), zone_info->nr_zones,
> +		emulated, zone_info->zone_size);
>  
>  	return 0;
Johannes Thumshirn Jan. 13, 2021, 6:23 p.m. UTC | #3
On 13/01/2021 19:00, David Sterba wrote:
> so what changed that it's fine to use device->fs_info now while it was
> not before?

The time when we call btrfs_get_dev_zone_info() in "btrfs: defer loading
zone info after opening trees". 

From the commit message:

"The current call site of btrfs_get_dev_zone_info() during the mount process
is earlier than reading the trees, so we can't slice a regular device to
conventional zones. This patch defers the loading of zone info to
open_ctree() to load the emulated zone size from a device extent."
    
Hope this clarifies your concerns.
diff mbox series

Patch

diff --git a/fs/btrfs/dev-replace.c b/fs/btrfs/dev-replace.c
index 324f646d6e5e..e77cb46bf15d 100644
--- a/fs/btrfs/dev-replace.c
+++ b/fs/btrfs/dev-replace.c
@@ -321,6 +321,9 @@  static int btrfs_init_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
 	set_blocksize(device->bdev, BTRFS_BDEV_BLOCKSIZE);
 	device->fs_devices = fs_info->fs_devices;
 
+	if (btrfs_is_zoned(fs_info) && bdev_zoned_model(bdev) == BLK_ZONED_NONE)
+		device->force_zoned = true;
+
 	ret = btrfs_get_dev_zone_info(device);
 	if (ret)
 		goto error;
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index 7d92b11ea603..2cdb5fe3e423 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -669,6 +669,15 @@  static int btrfs_open_one_device(struct btrfs_fs_devices *fs_devices,
 	clear_bit(BTRFS_DEV_STATE_IN_FS_METADATA, &device->dev_state);
 	device->mode = flags;
 
+	/* Emulate zoned mode on regular device? */
+	if ((btrfs_super_incompat_flags(disk_super) &
+	     BTRFS_FEATURE_INCOMPAT_ZONED) &&
+	    bdev_zoned_model(device->bdev) == BLK_ZONED_NONE) {
+		btrfs_info(NULL,
+"zoned: incompat zoned flag detected on regular device, forcing zoned mode emulation");
+		device->force_zoned = true;
+	}
+
 	fs_devices->open_devices++;
 	if (test_bit(BTRFS_DEV_STATE_WRITEABLE, &device->dev_state) &&
 	    device->devid != BTRFS_DEV_REPLACE_DEVID) {
@@ -2562,6 +2571,11 @@  int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path
 	device->fs_info = fs_info;
 	device->bdev = bdev;
 
+	/* Zoned mode is enabled. Emulate zoned device on a regular device. */
+	if (btrfs_is_zoned(fs_info) &&
+	    bdev_zoned_model(device->bdev) == BLK_ZONED_NONE)
+		device->force_zoned = true;
+
 	ret = btrfs_get_dev_zone_info(device);
 	if (ret)
 		goto error_free_device;
diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h
index 1997a4649a66..59d9d47f173d 100644
--- a/fs/btrfs/volumes.h
+++ b/fs/btrfs/volumes.h
@@ -144,6 +144,9 @@  struct btrfs_device {
 	struct completion kobj_unregister;
 	/* For sysfs/FSID/devinfo/devid/ */
 	struct kobject devid_kobj;
+
+	/* Force zoned mode */
+	bool force_zoned;
 };
 
 /*
diff --git a/fs/btrfs/zoned.c b/fs/btrfs/zoned.c
index ae566a7da088..fc43a650cd79 100644
--- a/fs/btrfs/zoned.c
+++ b/fs/btrfs/zoned.c
@@ -119,6 +119,32 @@  static inline u32 sb_zone_number(int shift, int mirror)
 	return 0;
 }
 
+static int emulate_report_zones(struct btrfs_device *device, u64 pos,
+				struct blk_zone *zones, unsigned int nr_zones)
+{
+	const sector_t zone_sectors =
+		device->fs_info->zone_size >> SECTOR_SHIFT;
+	sector_t bdev_size = device->bdev->bd_part->nr_sects;
+	unsigned int i;
+
+	pos >>= SECTOR_SHIFT;
+	for (i = 0; i < nr_zones; i++) {
+		zones[i].start = i * zone_sectors + pos;
+		zones[i].len = zone_sectors;
+		zones[i].capacity = zone_sectors;
+		zones[i].wp = zones[i].start + zone_sectors;
+		zones[i].type = BLK_ZONE_TYPE_CONVENTIONAL;
+		zones[i].cond = BLK_ZONE_COND_NOT_WP;
+
+		if (zones[i].wp >= bdev_size) {
+			i++;
+			break;
+		}
+	}
+
+	return i;
+}
+
 static int btrfs_get_dev_zones(struct btrfs_device *device, u64 pos,
 			       struct blk_zone *zones, unsigned int *nr_zones)
 {
@@ -127,6 +153,12 @@  static int btrfs_get_dev_zones(struct btrfs_device *device, u64 pos,
 	if (!*nr_zones)
 		return 0;
 
+	if (device->force_zoned) {
+		ret = emulate_report_zones(device, pos, zones, *nr_zones);
+		*nr_zones = ret;
+		return 0;
+	}
+
 	ret = blkdev_report_zones(device->bdev, pos >> SECTOR_SHIFT, *nr_zones,
 				  copy_zone_info_cb, zones);
 	if (ret < 0) {
@@ -143,6 +175,49 @@  static int btrfs_get_dev_zones(struct btrfs_device *device, u64 pos,
 	return 0;
 }
 
+static int calculate_emulated_zone_size(struct btrfs_fs_info *fs_info)
+{
+	struct btrfs_path *path;
+	struct btrfs_root *root = fs_info->dev_root;
+	struct btrfs_key key;
+	struct extent_buffer *leaf;
+	struct btrfs_dev_extent *dext;
+	int ret = 0;
+
+	key.objectid = 1;
+	key.type = BTRFS_DEV_EXTENT_KEY;
+	key.offset = 0;
+
+	path = btrfs_alloc_path();
+	if (!path)
+		return -ENOMEM;
+
+	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
+	if (ret < 0)
+		goto out;
+
+	if (path->slots[0] >= btrfs_header_nritems(path->nodes[0])) {
+		ret = btrfs_next_item(root, path);
+		if (ret < 0)
+			goto out;
+		/* No dev extents at all? Not good */
+		if (ret > 0) {
+			ret = -EUCLEAN;
+			goto out;
+		}
+	}
+
+	leaf = path->nodes[0];
+	dext = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_dev_extent);
+	fs_info->zone_size = btrfs_dev_extent_length(leaf, dext);
+	ret = 0;
+
+out:
+	btrfs_free_path(path);
+
+	return ret;
+}
+
 int btrfs_get_dev_zone_info_all_devices(struct btrfs_fs_info *fs_info)
 {
 	struct btrfs_fs_devices *fs_devices = fs_info->fs_devices;
@@ -158,6 +233,12 @@  int btrfs_get_dev_zone_info_all_devices(struct btrfs_fs_info *fs_info)
 		if (!device->bdev)
 			continue;
 
+		if (device->force_zoned && !fs_info->zone_size) {
+			ret = calculate_emulated_zone_size(fs_info);
+			if (ret)
+				break;
+		}
+
 		ret = btrfs_get_dev_zone_info(device);
 		if (ret)
 			break;
@@ -177,9 +258,11 @@  int btrfs_get_dev_zone_info(struct btrfs_device *device)
 	struct blk_zone *zones = NULL;
 	unsigned int i, nreported = 0, nr_zones;
 	unsigned int zone_sectors;
+	const bool force_zoned = device->force_zoned;
+	char *model, *emulated;
 	int ret;
 
-	if (!bdev_is_zoned(bdev))
+	if (!bdev_is_zoned(bdev) && !force_zoned)
 		return 0;
 
 	if (device->zone_info)
@@ -189,8 +272,12 @@  int btrfs_get_dev_zone_info(struct btrfs_device *device)
 	if (!zone_info)
 		return -ENOMEM;
 
+	if (force_zoned)
+		zone_sectors = device->fs_info->zone_size >> SECTOR_SHIFT;
+	else
+		zone_sectors = bdev_zone_sectors(bdev);
+
 	nr_sectors = bdev->bd_part->nr_sects;
-	zone_sectors = bdev_zone_sectors(bdev);
 	/* Check if it's power of 2 (see is_power_of_2) */
 	ASSERT(zone_sectors != 0 && (zone_sectors & (zone_sectors - 1)) == 0);
 	zone_info->zone_size = zone_sectors << SECTOR_SHIFT;
@@ -296,12 +383,22 @@  int btrfs_get_dev_zone_info(struct btrfs_device *device)
 
 	device->zone_info = zone_info;
 
-	/* device->fs_info is not safe to use for printing messages */
-	btrfs_info_in_rcu(NULL,
-			"host-%s zoned block device %s, %u zones of %llu bytes",
-			bdev_zoned_model(bdev) == BLK_ZONED_HM ? "managed" : "aware",
-			rcu_str_deref(device->name), zone_info->nr_zones,
-			zone_info->zone_size);
+	if (bdev_zoned_model(bdev) == BLK_ZONED_HM) {
+		model = "host-managed zoned";
+		emulated = "";
+	} else if (bdev_zoned_model(bdev) == BLK_ZONED_HA) {
+		model = "host-aware zoned";
+		emulated = "";
+	} else if (bdev_zoned_model(bdev) == BLK_ZONED_NONE &&
+		 device->force_zoned) {
+		model = "regular";
+		emulated = "emulated ";
+	}
+
+	btrfs_info_in_rcu(device->fs_info,
+		"%s block device %s, %u %szones of %llu bytes",
+		model, rcu_str_deref(device->name), zone_info->nr_zones,
+		emulated, zone_info->zone_size);
 
 	return 0;
 
@@ -348,7 +445,7 @@  int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info)
 	u64 nr_devices = 0;
 	u64 zone_size = 0;
 	u64 max_zone_append_size = 0;
-	const bool incompat_zoned = btrfs_is_zoned(fs_info);
+	const bool incompat_zoned = btrfs_fs_incompat(fs_info, ZONED);
 	int ret = 0;
 
 	/* Count zoned devices */
@@ -360,8 +457,10 @@  int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info)
 
 		model = bdev_zoned_model(device->bdev);
 		if (model == BLK_ZONED_HM ||
-		    (model == BLK_ZONED_HA && incompat_zoned)) {
-			struct btrfs_zoned_device_info *zone_info;
+		    (model == BLK_ZONED_HA && incompat_zoned) ||
+		    device->force_zoned) {
+			struct btrfs_zoned_device_info *zone_info =
+				device->zone_info;
 
 			zone_info = device->zone_info;
 			zoned_devices++;
diff --git a/fs/btrfs/zoned.h b/fs/btrfs/zoned.h
index 5e0e7de84a82..058a57317c05 100644
--- a/fs/btrfs/zoned.h
+++ b/fs/btrfs/zoned.h
@@ -143,12 +143,16 @@  static inline void btrfs_dev_clear_zone_empty(struct btrfs_device *device, u64 p
 static inline bool btrfs_check_device_zone_type(const struct btrfs_fs_info *fs_info,
 						struct block_device *bdev)
 {
-	u64 zone_size;
-
 	if (btrfs_is_zoned(fs_info)) {
-		zone_size = bdev_zone_sectors(bdev) << SECTOR_SHIFT;
-		/* Do not allow non-zoned device */
-		return bdev_is_zoned(bdev) && fs_info->zone_size == zone_size;
+		/*
+		 * We can allow a regular device on a zoned btrfs, because
+		 * we will emulate zoned device on the regular device.
+		 */
+		if (!bdev_is_zoned(bdev))
+			return true;
+
+		return fs_info->zone_size ==
+			(bdev_zone_sectors(bdev) << SECTOR_SHIFT);
 	}
 
 	/* Do not allow Host Manged zoned device */