Re: [PATCH v2 1/7] btrfs: do not call close_fs_devices in btrfs_rm_device

[Anand Jain <anand.jain@oracle.com>]

On 28/07/2021 05:01, Josef Bacik wrote:
> There's a subtle case where if we're removing the seed device from a
> file system we need to free its private copy of the fs_devices.  However
> we do not need to call close_fs_devices(), because at this point there
> are no devices left to close as we've closed the last one.  The only
> thing that close_fs_devices() does is decrement ->opened, which should
> be 1.  We want to avoid calling close_fs_devices() here because it has a
> lockdep_assert_held(&uuid_mutex), and we are going to stop holding the
> uuid_mutex in this path.
> 
> So add an assert for the ->opened counter and simply decrement it like
> we should, and then clean up like normal.  Also add a comment explaining
> what we're doing here as I initially removed this code erroneously.
> 
> Signed-off-by: Josef Bacik <josef@toxicpanda.com>
> ---
>   fs/btrfs/volumes.c | 10 +++++++++-
>   1 file changed, 9 insertions(+), 1 deletion(-)
> 
> diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
> index 86846d6e58d0..5217b93172b4 100644
> --- a/fs/btrfs/volumes.c
> +++ b/fs/btrfs/volumes.c
> @@ -2200,9 +2200,17 @@ int btrfs_rm_device(struct btrfs_fs_info *fs_info, const char *device_path,
>   	synchronize_rcu();
>   	btrfs_free_device(device);
>   
> +	/*
> +	 * This can happen if cur_devices is the private seed devices list.  We
> +	 * cannot call close_fs_devices() here because it expects the uuid_mutex
> +	 * to be held, but in fact we don't need that for the private
> +	 * seed_devices, we can simply decrement cur_devices->opened and then
> +	 * remove it from our list and free the fs_devices.
> +	 */

>   	if (cur_devices->open_devices == 0) {

  We should in fact use cur_devices->num_devices == 0 here.
  Sent a patch [1] to fix it.

[1]
https://patchwork.kernel.org/project/linux-btrfs/patch/d9c89b1740a876b3851fcf358f22809aa7f1ad2a.1630478246.git.anand.jain@oracle.com/


> +		ASSERT(cur_devices->opened == 1);

We don't need to assert(). free_fs_devices() has a warning for it.

         WARN_ON(fs_devices->opened);

>   		list_del_init(&cur_devices->seed_list);
> -		close_fs_devices(cur_devices);
> +		cur_devices->opened--;
>   		free_fs_devices(cur_devices);
>   	}


With the above two fixed.

Reviewed-by: Anand Jain <anand.jain@oracle.com>

Thanks, Anand