[SCSI] zfcp: No automatic port_rescan on events
In FC fabrics with large zones, the automatic port_rescan on incoming ELS
and any adapter recovery can cause quite some traffic at the very same
time, especially if lots of Linux images share an HBA, which is common on
s390. This can cause trouble and failures. Fix this by making such port
rescans dependent on a user configurable module parameter.
The following unconditional automatic port rescans remain as is:
On setting an adapter online and
on manual user-triggered writes to the sysfs attribute port_rescan.
Signed-off-by: Steffen Maier <maier@linux.vnet.ibm.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>
diff --git a/drivers/s390/scsi/zfcp_ccw.c b/drivers/s390/scsi/zfcp_ccw.c
index 9646766..f2dd3a0 100644
--- a/drivers/s390/scsi/zfcp_ccw.c
+++ b/drivers/s390/scsi/zfcp_ccw.c
@@ -57,7 +57,7 @@
zfcp_erp_adapter_reopen(adapter, ZFCP_STATUS_COMMON_ERP_FAILED,
tag);
zfcp_erp_wait(adapter);
- flush_work(&adapter->scan_work);
+ flush_work(&adapter->scan_work); /* ok to call even if nothing queued */
zfcp_ccw_adapter_put(adapter);
@@ -171,6 +171,11 @@
adapter->req_no = 0;
zfcp_ccw_activate(cdev, 0, "ccsonl1");
+ /* scan for remote ports
+ either at the end of any successful adapter recovery
+ or only after the adapter recovery for setting a device online */
+ zfcp_fc_inverse_conditional_port_scan(adapter);
+ flush_work(&adapter->scan_work); /* ok to call even if nothing queued */
zfcp_ccw_adapter_put(adapter);
return 0;
}