Re: [PATCH 2/2] writeback: use DEFINE_WAIT_BIT instead of DEFINE_WAIT for bit wait queue

[Jacob Wen <jian.w.wen@oracle.com>]

Can I have a review?

On 8/13/20 1:05 PM, Jacob Wen wrote:
> DEFINE_WAIT_BIT uses wake_bit_function() which is able to avoid
> false-wakeups due to possible hash collisions in the bit wait table.
>
> Signed-off-by: Jacob Wen <jian.w.wen@oracle.com>
> ---
>   fs/fs-writeback.c | 6 +++---
>   1 file changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c
> index a605c3dddabc..3bf751b33b48 100644
> --- a/fs/fs-writeback.c
> +++ b/fs/fs-writeback.c
> @@ -1354,16 +1354,16 @@ void inode_wait_for_writeback(struct inode *inode)
>   static void inode_sleep_on_writeback(struct inode *inode)
>   	__releases(inode->i_lock)
>   {
> -	DEFINE_WAIT(wait);
> +	DEFINE_WAIT_BIT(wait, &inode->i_state, __I_SYNC);
>   	wait_queue_head_t *wqh = bit_waitqueue(&inode->i_state, __I_SYNC);
>   	int sleep;
>   
> -	prepare_to_wait(wqh, &wait, TASK_UNINTERRUPTIBLE);
> +	prepare_to_wait(wqh, &wait.wq_entry, TASK_UNINTERRUPTIBLE);
>   	sleep = inode->i_state & I_SYNC;
>   	spin_unlock(&inode->i_lock);
>   	if (sleep)
>   		schedule();
> -	finish_wait(wqh, &wait);
> +	finish_wait(wqh, &wait.wq_entry);
>   }
>   
>   /*