Thread (10 messages) 10 messages, 2 authors, 2021-07-28

Re: [PATCH 1/6] btrfs: do not check for ->num_devices == 0 in rm_device

From: Anand Jain <hidden>
Date: 2021-07-27 22:35:39

On 28/07/2021 03:47, Josef Bacik wrote:
We're specifically prohibited from removing the last device in a file
system, so this check is meaningless and the code can be deleted.
Josef, That's not true when removing a seed device from the sprouted 
filesystem.
quoted hunk ↗ jump to hunk
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
---
  fs/btrfs/volumes.c | 7 -------
  1 file changed, 7 deletions(-)
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index 86846d6e58d0..373be4e54f28 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -2199,13 +2199,6 @@ int btrfs_rm_device(struct btrfs_fs_info *fs_info, const char *device_path,
  	btrfs_close_bdev(device);
  	synchronize_rcu();
  	btrfs_free_device(device);
-
-	if (cur_devices->open_devices == 0) {
-		list_del_init(&cur_devices->seed_list);
-		close_fs_devices(cur_devices);
-		free_fs_devices(cur_devices);
-	}
-
  Here the cur_devices will point to the seed devices, and generally, it 
is the last opened devices in its fs_devices structure.

  Comments about it is a few lines above from this.

2157 /*
2158 * In normal cases the cur_devices == fs_devices. But in case
2159 * of deleting a seed device, the cur_devices should point to
2160 * its own fs_devices listed under the fs_devices->seed.
2161 */

Thanks, Anand
  out:
  	mutex_unlock(&uuid_mutex);
  	return ret;
  
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help