Re: [PATCH v6 07/13] Fix race between starved list processing and device removal

[Tejun Heo <tj@kernel.org>]

Hello, Bart.

On Wed, Nov 28, 2012 at 01:48:24PM +0100, Bart Van Assche wrote:
> @@ -452,11 +452,13 @@ static void scsi_run_queue(struct request_queue *q)
>  			continue;
>  		}
>  
> -		spin_unlock(shost->host_lock);
> -		spin_lock(sdev->request_queue->queue_lock);
> -		__blk_run_queue(sdev->request_queue);
> -		spin_unlock(sdev->request_queue->queue_lock);
> -		spin_lock(shost->host_lock);
> +		get_device(&sdev->sdev_gendev);
> +		spin_unlock_irqrestore(shost->host_lock, flags);
> +
> +		blk_run_queue(sdev->request_queue);
> +		put_device(&sdev->sdev_gendev);
> +
> +		spin_lock_irqsave(shost->host_lock, flags);

Some comment wouldn't hurt here.

> @@ -977,6 +978,10 @@ void __scsi_remove_device(struct scsi_device *sdev)
>  	blk_cleanup_queue(sdev->request_queue);
>  	cancel_work_sync(&sdev->requeue_work);
>  
> +	spin_lock_irqsave(shost->host_lock, flags);
> +	list_del(&sdev->starved_entry);
> +	spin_unlock_irqrestore(shost->host_lock, flags);
> +

And please add a comment explaining what's going on around
->starved_entry.  It's rather different from anything else.

Thanks!

-- 
tejun