Re: [PATCH -mm] libata-core: convert to use cancel_rearming_delayed_work()

[Tejun Heo]

Hello,

Oleg Nesterov wrote:
>  void ata_port_queue_task(struct ata_port *ap, work_func_t fn, void *data,
>  			 unsigned long delay)
>  {
> -	int rc;
> -
> -	if (ap->pflags & ATA_PFLAG_FLUSH_PORT_TASK)
> -		return;
> -
>  	PREPARE_DELAYED_WORK(&ap->port_task, fn);
>  	ap->port_task_data = data;
>  
> -	rc = queue_delayed_work(ata_wq, &ap->port_task, delay);
> -
> -	/* rc == 0 means that another user is using port task */
> -	WARN_ON(rc == 0);
> +	/* may fail if ata_port_flush_task() in progress */
> +	queue_delayed_work(ata_wq, &ap->port_task, delay);
>  }

I don't really think ata_port_queue_task() deserves to live after this.
 Feel free to kill it.

>  void ata_port_flush_task(struct ata_port *ap)
>  {
> -	unsigned long flags;
> -
>  	DPRINTK("ENTER\n");
>  
> -	spin_lock_irqsave(ap->lock, flags);
> -	ap->pflags |= ATA_PFLAG_FLUSH_PORT_TASK;
> -	spin_unlock_irqrestore(ap->lock, flags);
> -
> -	DPRINTK("flush #1\n");
> -	cancel_work_sync(&ap->port_task.work); /* akpm: seems unneeded */
> -
> -	/*
> -	 * At this point, if a task is running, it's guaranteed to see
> -	 * the FLUSH flag; thus, it will never queue pio tasks again.
> -	 * Cancel and flush.
> -	 */
> -	if (!cancel_delayed_work(&ap->port_task)) {
> -		if (ata_msg_ctl(ap))
> -			ata_port_printk(ap, KERN_DEBUG, "%s: flush #2\n",
> -					__FUNCTION__);
> -		cancel_work_sync(&ap->port_task.work);
> -	}
> -
> -	spin_lock_irqsave(ap->lock, flags);
> -	ap->pflags &= ~ATA_PFLAG_FLUSH_PORT_TASK;
> -	spin_unlock_irqrestore(ap->lock, flags);
> +	cancel_rearming_delayed_work(&ap->port_task);

Nor does ata_port_flush_task().

> @@ -6094,13 +6064,7 @@ void ata_port_detach(struct ata_port *ap
>  	spin_unlock_irqrestore(ap->lock, flags);
>  
>  	ata_port_wait_eh(ap);
> -
> -	/* Flush hotplug task.  The sequence is similar to
> -	 * ata_port_flush_task().
> -	 */
> -	cancel_work_sync(&ap->hotplug_task.work); /* akpm: why? */
> -	cancel_delayed_work(&ap->hotplug_task);
> -	cancel_work_sync(&ap->hotplug_task.work);
> +	cancel_rearming_delayed_work(&ap->hotplug_task);

Yay, I've always wanted to do that.  :-)

Thanks.

-- 
tejun
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/