[SCSI] qla2xxx: Correct abort issue during loop-down state.

Correct issue where abort I/O command was not being issued
when the loop-state was down.
Signed-off-by: NRavi Anand <ravi.anand@qlogic.com>
Signed-off-by: NAndrew Vasquez <andrew.vasquez@qlogic.com>
Signed-off-by: NJames Bottomley <James.Bottomley@SteelEye.com>

drivers/scsi/qla2xxx/qla_mbx.c

@@ -868,10 +868,6 @@ qla2x00_abort_command(scsi_qla_host_t *ha, srb_t *sp)
 	DEBUG11(printk("qla2x00_abort_command(%ld): entered.\n", ha->host_no);)
 
 	fcport = sp->fcport;
-	if (atomic_read(&ha->loop_state) == LOOP_DOWN ||
-	    atomic_read(&fcport->state) == FCS_DEVICE_LOST) {
-		return 1;
-	}
 
 	spin_lock_irqsave(&ha->hardware_lock, flags);
 	for (handle = 1; handle < MAX_OUTSTANDING_COMMANDS; handle++) {
@@ -2181,10 +2177,6 @@ qla24xx_abort_command(scsi_qla_host_t *ha, srb_t *sp)
 	DEBUG11(printk("%s(%ld): entered.\n", __func__, ha->host_no);)
 
 	fcport = sp->fcport;
-	if (atomic_read(&ha->loop_state) == LOOP_DOWN ||
-	    atomic_read(&fcport->state) == FCS_DEVICE_LOST) {
-		return QLA_FUNCTION_FAILED;
-	}
 
 	spin_lock_irqsave(&ha->hardware_lock, flags);
 	for (handle = 1; handle < MAX_OUTSTANDING_COMMANDS; handle++) {