On 2020-07-01 18:32, Stanley Chu wrote:
> diff --git a/drivers/scsi/ufs/ufshcd.c b/drivers/scsi/ufs/ufshcd.c
> index 59358bb75014..cadfa9006972 100644
> --- a/drivers/scsi/ufs/ufshcd.c
> +++ b/drivers/scsi/ufs/ufshcd.c
> @@ -8599,10 +8599,14 @@ EXPORT_SYMBOL(ufshcd_runtime_idle);
> int ufshcd_shutdown(struct ufs_hba *hba)
> {
> int ret = 0;
> + struct scsi_target *starget;
>
> if (!hba->is_powered)
> goto out;
>
> + list_for_each_entry(starget, &hba->host->__targets, siblings)
> + scsi_target_quiesce(starget);
> +
> if (ufshcd_is_ufs_dev_poweroff(hba) && ufshcd_is_link_off(hba))
> goto out;
Please add a comment above the list_for_each_entry() loop that explains
that there is no matching scsi_target_unquiesce() call and also that
SCSI commands queued after the scsi_target_quiesce() call returned will
block until blk_cleanup_queue() is called (see also the blk_queue_dying()
check in blk_queue_enter()).
Thanks,
Bart.
