[SCSI] pm8001: deficient responses to IO_XFER_ERROR_BREAK and IO_XFER_OPEN_RETRY_TIMEOUT

IO_XFER_ERROR_BREAK and IO_XFER_OPEN_RETRY_TIMEOUT are deficient of the
required actions as outlined in the programming manual for the pm8001. Due to
the overlapping code requirements of these recovery responses, we found it
necessary to bundle them together into one patch.

When a break is received during the command phase (ssp_completion), this is a
result of a timeout or interruption on the bus. Logic suggests that we should
retry the command.

When a break is received during the data-phase (ssp_event), the task must be
aborted on the target or it will retain a data-phase lock turning the target
reticent to all future media commands yet will successfully respond to TUR,
INQUIRY and ABORT leading eventually to target failure through several
abort-cycle loops.

The open retry interval is exceedingly short resulting in occasional target
drop-off during expander resets or when targets push-back during bad-block
remapping. Increased effective timeout from 130ms to 1.5 seconds for each try
so as to trigger after the administrative inquiry/tur timeout in the scsi
subsystem to keep error-recovery harmonics to a minimum.

When an open retry timeout event is received, the action required by the
targets is to issue an abort for the outstanding command then logic suggests
we retry the command as this state is usually an indication of a credit block
or busy condition on the target.

We hijacked the pm8001_handle_event work queue handler so that it will handle
task as an argument instead of device for the workers in support of the
deferred handling outlined above.

Moderate to Heavy bad-path testing on a 2.6.32 vintage kernel, compile-testing
on scsi-misc-2.6 kernel ...

Signed-off-by: Mark Salyzyn <mark_salyzyn@xyratex.com>
Acked-by: Jack Wang <jack_wang@usish.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>

1 files changed, 72 insertions, 0 deletions

diff --git a/drivers/scsi/pm8001/pm8001_sas.c b/drivers/scsi/pm8001/pm8001_sas.c
index ab0704e39040..9589fc941a8b 100644
--- a/drivers/scsi/pm8001/pm8001_sas.c
+++ b/drivers/scsi/pm8001/pm8001_sas.c
@@ -538,6 +538,7 @@ void pm8001_ccb_task_free(struct pm8001_hba_info *pm8001_ha,
 	task->lldd_task = NULL;
 	ccb->task = NULL;
 	ccb->ccb_tag = 0xFFFFFFFF;
+	ccb->open_retry = 0;
 	pm8001_ccb_free(pm8001_ha, ccb_idx);
 }
 
@@ -882,6 +883,77 @@ static int pm8001_issue_ssp_tmf(struct domain_device *dev,
 		tmf);
 }
 
+/* retry commands by ha, by task and/or by device */
+void pm8001_open_reject_retry(
+	struct pm8001_hba_info *pm8001_ha,
+	struct sas_task *task_to_close,
+	struct pm8001_device *device_to_close)
+{
+	int i;
+	unsigned long flags;
+
+	if (pm8001_ha == NULL)
+		return;
+
+	spin_lock_irqsave(&pm8001_ha->lock, flags);
+
+	for (i = 0; i < PM8001_MAX_CCB; i++) {
+		struct sas_task *task;
+		struct task_status_struct *ts;
+		struct pm8001_device *pm8001_dev;
+		unsigned long flags1;
+		u32 tag;
+		struct pm8001_ccb_info *ccb = &pm8001_ha->ccb_info[i];
+
+		pm8001_dev = ccb->device;
+		if (!pm8001_dev || (pm8001_dev->dev_type == NO_DEVICE))
+			continue;
+		if (!device_to_close) {
+			uintptr_t d = (uintptr_t)pm8001_dev
+					- (uintptr_t)&pm8001_ha->devices;
+			if (((d % sizeof(*pm8001_dev)) != 0)
+			 || ((d / sizeof(*pm8001_dev)) >= PM8001_MAX_DEVICES))
+				continue;
+		} else if (pm8001_dev != device_to_close)
+			continue;
+		tag = ccb->ccb_tag;
+		if (!tag || (tag == 0xFFFFFFFF))
+			continue;
+		task = ccb->task;
+		if (!task || !task->task_done)
+			continue;
+		if (task_to_close && (task != task_to_close))
+			continue;
+		ts = &task->task_status;
+		ts->resp = SAS_TASK_COMPLETE;
+		/* Force the midlayer to retry */
+		ts->stat = SAS_OPEN_REJECT;
+		ts->open_rej_reason = SAS_OREJ_RSVD_RETRY;
+		if (pm8001_dev)
+			pm8001_dev->running_req--;
+		spin_lock_irqsave(&task->task_state_lock, flags1);
+		task->task_state_flags &= ~SAS_TASK_STATE_PENDING;
+		task->task_state_flags &= ~SAS_TASK_AT_INITIATOR;
+		task->task_state_flags |= SAS_TASK_STATE_DONE;
+		if (unlikely((task->task_state_flags
+				& SAS_TASK_STATE_ABORTED))) {
+			spin_unlock_irqrestore(&task->task_state_lock,
+				flags1);
+			pm8001_ccb_task_free(pm8001_ha, task, ccb, tag);
+		} else {
+			spin_unlock_irqrestore(&task->task_state_lock,
+				flags1);
+			pm8001_ccb_task_free(pm8001_ha, task, ccb, tag);
+			mb();/* in order to force CPU ordering */
+			spin_unlock_irqrestore(&pm8001_ha->lock, flags);
+			task->task_done(task);
+			spin_lock_irqsave(&pm8001_ha->lock, flags);
+		}
+	}
+
+	spin_unlock_irqrestore(&pm8001_ha->lock, flags);
+}
+
 /**
   * Standard mandates link reset for ATA  (type 0) and hard reset for
   * SSP (type 1) , only for RECOVERY