s390/cio: recover from bad paths

In some situations we don't receive notification from firmware that
a previously unusable channelpath is usable again.

Schedule recovery for devices that return from path verification
without using all potentially usable paths. The recovery thread will
periodically trigger a path verification on the affected devices.

Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Suggested-by: Peter Oberparleiter <oberpar@linux.vnet.ibm.com>
Reviewed-by: Peter Oberparleiter <oberpar@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>

4 files changed, 24 insertions, 3 deletions

diff --git a/drivers/s390/cio/device.c b/drivers/s390/cio/device.c
index 489b583f263d..e5c32f4b5287 100644
--- a/drivers/s390/cio/device.c
+++ b/drivers/s390/cio/device.c
@@ -1225,10 +1225,16 @@ static int device_is_disconnected(struct ccw_device *cdev)
 static int recovery_check(struct device *dev, void *data)
 {
 	struct ccw_device *cdev = to_ccwdev(dev);
+	struct subchannel *sch;
 	int *redo = data;
 
 	spin_lock_irq(cdev->ccwlock);
 	switch (cdev->private->state) {
+	case DEV_STATE_ONLINE:
+		sch = to_subchannel(cdev->dev.parent);
+		if ((sch->schib.pmcw.pam & sch->opm) == sch->vpm)
+			break;
+		/* fall through */
 	case DEV_STATE_DISCONNECTED:
 		CIO_MSG_EVENT(3, "recovery: trigger 0.%x.%04x\n",
 			      cdev->private->dev_id.ssid,
@@ -1260,7 +1266,7 @@ static void recovery_work_func(struct work_struct *unused)
 		}
 		spin_unlock_irq(&recovery_lock);
 	} else
-		CIO_MSG_EVENT(4, "recovery: end\n");
+		CIO_MSG_EVENT(3, "recovery: end\n");
 }
 
 static DECLARE_WORK(recovery_work, recovery_work_func);
@@ -1274,11 +1280,11 @@ static void recovery_func(unsigned long data)
 	schedule_work(&recovery_work);
 }
 
-static void ccw_device_schedule_recovery(void)
+void ccw_device_schedule_recovery(void)
 {
 	unsigned long flags;
 
-	CIO_MSG_EVENT(4, "recovery: schedule\n");
+	CIO_MSG_EVENT(3, "recovery: schedule\n");
 	spin_lock_irqsave(&recovery_lock, flags);
 	if (!timer_pending(&recovery_timer) || (recovery_phase != 0)) {
 		recovery_phase = 0;
diff --git a/drivers/s390/cio/device.h b/drivers/s390/cio/device.h
index ec497af99dd8..69cb70f080a5 100644
--- a/drivers/s390/cio/device.h
+++ b/drivers/s390/cio/device.h
@@ -134,6 +134,7 @@ void ccw_device_set_disconnected(struct ccw_device *cdev);
 void ccw_device_set_notoper(struct ccw_device *cdev);
 
 void ccw_device_set_timeout(struct ccw_device *, int);
+void ccw_device_schedule_recovery(void);
 
 /* Channel measurement facility related */
 void retry_set_schib(struct ccw_device *cdev);
diff --git a/drivers/s390/cio/device_fsm.c b/drivers/s390/cio/device_fsm.c
index 12016e32e519..f98ea674c3d8 100644
--- a/drivers/s390/cio/device_fsm.c
+++ b/drivers/s390/cio/device_fsm.c
@@ -476,6 +476,17 @@ static void create_fake_irb(struct irb *irb, int type)
 	}
 }
 
+static void ccw_device_handle_broken_paths(struct ccw_device *cdev)
+{
+	struct subchannel *sch = to_subchannel(cdev->dev.parent);
+	u8 broken_paths = (sch->schib.pmcw.pam & sch->opm) ^ sch->vpm;
+
+	if (broken_paths && (cdev->private->path_broken_mask != broken_paths))
+		ccw_device_schedule_recovery();
+
+	cdev->private->path_broken_mask = broken_paths;
+}
+
 void ccw_device_verify_done(struct ccw_device *cdev, int err)
 {
 	struct subchannel *sch;
@@ -508,6 +519,7 @@ callback:
 			memset(&cdev->private->irb, 0, sizeof(struct irb));
 		}
 		ccw_device_report_path_events(cdev);
+		ccw_device_handle_broken_paths(cdev);
 		break;
 	case -ETIME:
 	case -EUSERS:
diff --git a/drivers/s390/cio/io_sch.h b/drivers/s390/cio/io_sch.h
index 220f49145b2f..9a1b56b2df3e 100644
--- a/drivers/s390/cio/io_sch.h
+++ b/drivers/s390/cio/io_sch.h
@@ -131,6 +131,8 @@ struct ccw_device_private {
 				   not operable */
 	u8 path_gone_mask;	/* mask of paths, that became unavailable */
 	u8 path_new_mask;	/* mask of paths, that became available */
+	u8 path_broken_mask;	/* mask of paths, which were found to be
+				   unusable */
 	struct {
 		unsigned int fast:1;	/* post with "channel end" */
 		unsigned int repall:1;	/* report every interrupt status */