scsi: smartpqi: fix LUN reset when fw bkgnd thread is hung

Add support for a timeout on LUN resets.

Link: https://lore.kernel.org/r/157048750055.11757.9689400788261610618.stgit@brunhilda
Reviewed-by: Scott Benesh <scott.benesh@microsemi.com>
Reviewed-by: Scott Teel <scott.teel@microsemi.com>
Reviewed-by: Kevin Barnett <kevin.barnett@microsemi.com>
Signed-off-by: Murthy Bhat <Murthy.Bhat@microsemi.com>
Signed-off-by: Don Brace <don.brace@microsemi.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

2 files changed, 22 insertions, 4 deletions

diff --git a/drivers/scsi/smartpqi/smartpqi.h b/drivers/scsi/smartpqi/smartpqi.h
index af0662b8a09c..c09d48366d38 100644
--- a/drivers/scsi/smartpqi/smartpqi.h
+++ b/drivers/scsi/smartpqi/smartpqi.h
@@ -387,7 +387,8 @@ struct pqi_task_management_request {
 	struct pqi_iu_header header;
 	__le16	request_id;
 	__le16	nexus_id;
-	u8	reserved[4];
+	u8	reserved[2];
+	__le16  timeout;
 	u8	lun_number[8];
 	__le16	protocol_specific;
 	__le16	outbound_queue_id_to_manage;
@@ -764,6 +765,7 @@ struct pqi_config_table_firmware_features {
 #define PQI_FIRMWARE_FEATURE_SMP			1
 #define PQI_FIRMWARE_FEATURE_SOFT_RESET_HANDSHAKE	11
 #define PQI_FIRMWARE_FEATURE_RAID_IU_TIMEOUT		13
+#define PQI_FIRMWARE_FEATURE_TMF_IU_TIMEOUT		14
 
 struct pqi_config_table_debug {
 	struct pqi_config_table_section_header header;
@@ -1142,6 +1144,7 @@ struct pqi_ctrl_info {
 	u8		pqi_reset_quiesce_supported : 1;
 	u8		soft_reset_handshake_supported : 1;
 	u8		raid_iu_timeout_supported: 1;
+	u8		tmf_iu_timeout_supported: 1;
 
 	struct list_head scsi_device_list;
 	spinlock_t	scsi_device_list_lock;
diff --git a/drivers/scsi/smartpqi/smartpqi_init.c b/drivers/scsi/smartpqi/smartpqi_init.c
index 9ae17bde0b30..549455f3f4ae 100644
--- a/drivers/scsi/smartpqi/smartpqi_init.c
+++ b/drivers/scsi/smartpqi/smartpqi_init.c
@@ -5690,7 +5690,8 @@ static void pqi_lun_reset_complete(struct pqi_io_request *io_request,
 	complete(waiting);
 }
 
-#define PQI_LUN_RESET_TIMEOUT_SECS	10
+#define PQI_LUN_RESET_TIMEOUT_SECS	60
+#define PQI_LUN_RESET_POLL_COMPLETION_SECS	10
 
 static int pqi_wait_for_lun_reset_completion(struct pqi_ctrl_info *ctrl_info,
 	struct pqi_scsi_dev *device, struct completion *wait)
@@ -5699,7 +5700,7 @@ static int pqi_wait_for_lun_reset_completion(struct pqi_ctrl_info *ctrl_info,
 
 	while (1) {
 		if (wait_for_completion_io_timeout(wait,
-			PQI_LUN_RESET_TIMEOUT_SECS * PQI_HZ)) {
+			PQI_LUN_RESET_POLL_COMPLETION_SECS * PQI_HZ)) {
 			rc = 0;
 			break;
 		}
@@ -5736,6 +5737,9 @@ static int pqi_lun_reset(struct pqi_ctrl_info *ctrl_info,
 	memcpy(request->lun_number, device->scsi3addr,
 		sizeof(request->lun_number));
 	request->task_management_function = SOP_TASK_MANAGEMENT_LUN_RESET;
+	if (ctrl_info->tmf_iu_timeout_supported)
+		put_unaligned_le16(PQI_LUN_RESET_TIMEOUT_SECS,
+					&request->timeout);
 
 	pqi_start_io(ctrl_info,
 		&ctrl_info->queue_groups[PQI_DEFAULT_QUEUE_GROUP], RAID_PATH,
@@ -5765,7 +5769,7 @@ static int _pqi_device_reset(struct pqi_ctrl_info *ctrl_info,
 
 	for (retries = 0;;) {
 		rc = pqi_lun_reset(ctrl_info, device);
-		if (rc != -EAGAIN || ++retries > PQI_LUN_RESET_RETRIES)
+		if (rc == 0 || ++retries > PQI_LUN_RESET_RETRIES)
 			break;
 		msleep(PQI_LUN_RESET_RETRY_INTERVAL_MSECS);
 	}
@@ -6886,6 +6890,10 @@ static void pqi_ctrl_update_feature_flags(struct pqi_ctrl_info *ctrl_info,
 		ctrl_info->raid_iu_timeout_supported =
 			firmware_feature->enabled;
 		break;
+	case PQI_FIRMWARE_FEATURE_TMF_IU_TIMEOUT:
+		ctrl_info->tmf_iu_timeout_supported =
+			firmware_feature->enabled;
+		break;
 	}
 
 	pqi_firmware_feature_status(ctrl_info, firmware_feature);
@@ -6921,6 +6929,11 @@ static struct pqi_firmware_feature pqi_firmware_features[] = {
 		.feature_bit = PQI_FIRMWARE_FEATURE_RAID_IU_TIMEOUT,
 		.feature_status = pqi_ctrl_update_feature_flags,
 	},
+	{
+		.feature_name = "TMF IU Timeout",
+		.feature_bit = PQI_FIRMWARE_FEATURE_TMF_IU_TIMEOUT,
+		.feature_status = pqi_ctrl_update_feature_flags,
+	},
 };
 
 static void pqi_process_firmware_features(
@@ -8941,6 +8954,8 @@ static void __attribute__((unused)) verify_structures(void)
 	BUILD_BUG_ON(offsetof(struct pqi_task_management_request,
 		nexus_id) != 10);
 	BUILD_BUG_ON(offsetof(struct pqi_task_management_request,
+		timeout) != 14);
+	BUILD_BUG_ON(offsetof(struct pqi_task_management_request,
 		lun_number) != 16);
 	BUILD_BUG_ON(offsetof(struct pqi_task_management_request,
 		protocol_specific) != 24);