[PATCH] btrfs: zoned: fix btrfs_can_activate_zone() to support DUP profile

[PATCH] btrfs: zoned: fix btrfs_can_activate_zone() to support DUP profile

[Naohiro Aota]

btrfs_can_activate_zone() returns true if at least one device has one zone
available for activation. This is OK for the single profile, but not OK for
DUP profile. We need two zones to create a DUP block group. Fix it buy
properly handle the case with the profile flags.

Fixes: 265f7237dd25 ("btrfs: zoned: allow DUP on meta-data block groups")
CC: stable@vger.kernel.org # 6.1+
Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
---
 fs/btrfs/zoned.c | 14 ++++++++++++--
 1 file changed, 12 insertions(+), 2 deletions(-)

diff --git a/fs/btrfs/zoned.c b/fs/btrfs/zoned.c
index f3fcc3e09550..f7397680cde9 100644
--- a/fs/btrfs/zoned.c
+++ b/fs/btrfs/zoned.c
@@ -2099,11 +2099,21 @@ bool btrfs_can_activate_zone(struct btrfs_fs_devices *fs_devices, u64 flags)
 		if (!device->bdev)
 			continue;
 
-		if (!zinfo->max_active_zones ||
-		    atomic_read(&zinfo->active_zones_left)) {
+		if (!zinfo->max_active_zones) {
 			ret = true;
 			break;
 		}
+
+		switch (flags & BTRFS_BLOCK_GROUP_PROFILE_MASK) {
+		case 0: /* single */
+			ret = atomic_read(&zinfo->active_zones_left) >= 1;
+			break;
+		case BTRFS_BLOCK_GROUP_DUP:
+			ret = atomic_read(&zinfo->active_zones_left) >= 2;
+			break;
+		}
+		if (ret)
+			break;
 	}
 	mutex_unlock(&fs_info->chunk_mutex);
 
-- 
2.39.2

Re: [PATCH] btrfs: zoned: fix btrfs_can_activate_zone() to support DUP profile

[Johannes Thumshirn]

Looks good,
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>

Re: [PATCH] btrfs: zoned: fix btrfs_can_activate_zone() to support DUP profile

[David Sterba]

On Mon, Mar 13, 2023 at 04:29:49PM +0900, Naohiro Aota wrote:
> btrfs_can_activate_zone() returns true if at least one device has one zone
> available for activation. This is OK for the single profile, but not OK for
> DUP profile. We need two zones to create a DUP block group. Fix it buy
> properly handle the case with the profile flags.
> 
> Fixes: 265f7237dd25 ("btrfs: zoned: allow DUP on meta-data block groups")
> CC: stable@vger.kernel.org # 6.1+
> Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>

Added to misc-next, thanks.

> ---
>  fs/btrfs/zoned.c | 14 ++++++++++++--
>  1 file changed, 12 insertions(+), 2 deletions(-)
> 
> diff --git a/fs/btrfs/zoned.c b/fs/btrfs/zoned.c
> index f3fcc3e09550..f7397680cde9 100644
> --- a/fs/btrfs/zoned.c
> +++ b/fs/btrfs/zoned.c
> @@ -2099,11 +2099,21 @@ bool btrfs_can_activate_zone(struct btrfs_fs_devices *fs_devices, u64 flags)
>  		if (!device->bdev)
>  			continue;
>  
> -		if (!zinfo->max_active_zones ||
> -		    atomic_read(&zinfo->active_zones_left)) {
> +		if (!zinfo->max_active_zones) {
>  			ret = true;
>  			break;
>  		}
> +
> +		switch (flags & BTRFS_BLOCK_GROUP_PROFILE_MASK) {
> +		case 0: /* single */
> +			ret = atomic_read(&zinfo->active_zones_left) >= 1;

With ( ) around the expression it's more obvious that it's not a simple
assignment:

			ret = (atomic_read(&zinfo->active_zones_left) >= 1);

Updated in the commit.