Re: [PATCH v6] sbitmap: fix io hung due to race on sbitmap_word::cleared

[Bart Van Assche]

On 7/11/24 6:39 AM, Ming Lei wrote:
> There are only two WRITE on 'cleared':
>
> - xchg(&map->cleared, 0) in sbitmap_deferred_clear()
>
> - set_bit() in sbitmap_deferred_clear_bit()
>
> xchg() supposes to provide such protection already.

Hi Ming,

The comment above 'swap_lock' in this patch is as follows:

	/**
	 * @swap_lock: Held while swapping word <-> cleared
	 */

In other words, 'swap_lock' is used to serialize *code*. Using
synchronization objects to serialize code is known as an anti-pattern,
something that shouldn't be done. Synchronization objects should be used
to serialize access to data. Hence my question whether it would be
appropriate to protect all 'cleared' changes with the newly introduced
spinlock.

Thanks,

Bart.