Re: [RFC 7/9] zfcp: use fc_block_rport for TMFs and host reset to decouple from scsi_cmnd

[Hannes Reinecke <hare@suse.de>]

On 07/25/2017 04:14 PM, Steffen Maier wrote:
> Signed-off-by: Steffen Maier <maier@linux.vnet.ibm.com>
> ---
>  drivers/s390/scsi/zfcp_scsi.c | 6 ++++--
>  1 file changed, 4 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/s390/scsi/zfcp_scsi.c b/drivers/s390/scsi/zfcp_scsi.c
> index 05c823ccb959..8e96196fa877 100644
> --- a/drivers/s390/scsi/zfcp_scsi.c
> +++ b/drivers/s390/scsi/zfcp_scsi.c
> @@ -287,7 +287,7 @@ static int zfcp_task_mgmt_function(struct scsi_cmnd *scpnt, u8 tm_flags)
>  			break;
>  
>  		zfcp_erp_wait(adapter);
> -		ret = fc_block_scsi_eh(scpnt);
> +		ret = port->rport ? fc_block_rport(port->rport) : 0;
>  		if (ret) {
>  			zfcp_dbf_scsi_devreset("fiof", adapter, tm_flags, NULL,
>  					       scsi_id, scsi_lun);> @@ -337,11 +337,13 @@ static int
zfcp_scsi_eh_host_reset_handler(struct scsi_cmnd *scpnt)
>  {
>  	struct zfcp_scsi_dev *zfcp_sdev = sdev_to_zfcp(scpnt->device);
>  	struct zfcp_adapter *adapter = zfcp_sdev->port->adapter;
> +	struct zfcp_port *port;
>  	int ret;
>  
>  	zfcp_erp_adapter_reopen(adapter, 0, "schrh_1");
>  	zfcp_erp_wait(adapter);
> -	ret = fc_block_scsi_eh(scpnt);
> +	port = zfcp_sdev->port;
> +	ret = port->rport ? fc_block_rport(port->rport) : 0;
>  	if (ret)
>  		return ret;
>  
> 

Hmm. So there is a chance where ->rport is simply NULL, in which case we
won't be calling fc_block_rport().
However, we _do_ continue with TMF even then.
Wasn't that precisely the point of fc_block_rport() to figure out if the
rport is valid?
So shouldn't we rather return FAILED or something here?
If not, why do we continue sending TMFs to a non-existing port?

Cheers,

Hannes
-- 
Dr. Hannes Reinecke		   Teamlead Storage & Networking
hare@suse.de			               +49 911 74053 688
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: F. Imendörffer, J. Smithard, J. Guild, D. Upmanyu, G. Norton
HRB 21284 (AG Nürnberg)