Re: [PATCH 1/2] block: loop: share code of reread partitions

[Christoph Hellwig <hch@lst.de>]

On Tue, Jul 07, 2020 at 04:45:51PM +0800, Ming Lei wrote:
> loop_reread_partitions() has been there for rereading partitions, so
> replace the open code in __loop_clr_fd() with loop_reread_partitions()
> by passing 'locked' parameter.
> 
> Signed-off-by: Ming Lei <ming.lei@redhat.com>
> ---
>  drivers/block/loop.c | 29 ++++++++++++-----------------
>  1 file changed, 12 insertions(+), 17 deletions(-)
> 
> diff --git a/drivers/block/loop.c b/drivers/block/loop.c
> index a943207705dd..0e08468b9ce0 100644
> --- a/drivers/block/loop.c
> +++ b/drivers/block/loop.c
> @@ -650,13 +650,17 @@ static inline void loop_update_dio(struct loop_device *lo)
>  }
>  
>  static void loop_reread_partitions(struct loop_device *lo,
> -				   struct block_device *bdev)
> +				   struct block_device *bdev, bool locked)
>  {
>  	int rc;
>  
> -	mutex_lock(&bdev->bd_mutex);
> -	rc = bdev_disk_changed(bdev, false);
> -	mutex_unlock(&bdev->bd_mutex);
> +	if (locked) {
> +		rc = bdev_disk_changed(bdev, false);
> +	} else {
> +		mutex_lock(&bdev->bd_mutex);
> +		rc = bdev_disk_changed(bdev, false);
> +		mutex_unlock(&bdev->bd_mutex);
> +	}

functions with an argument based locking context are a really bad
idea.  And there is absolutely no reason to add them just for
a shared printk.