isci: atomic device lookup and reference counting

We have unsafe references to remote devices that are notified to
disappear at lldd_dev_gone.  In order to clean this up we need a single
canonical source for device lookups and stable references once a lookup
succeeds.  Towards that end guarantee that domain_device.lldd_dev is
NULL as soon as we start the process of stopping a device.  Any code
path that wants to safely lookup a remote device must do so through
task->dev->lldd_dev (isci_lookup_device()).

For in-flight references outside of scic_lock we need reference counting
to ensure that the device is not recycled before we are done with it.
Simplify device back references to just scic_sds_request.target_device
which is now the only permissible internal reference that is maintained
relative to the reference count.

There were two occasions where we wanted new i/o's to be treated as
SAS_TASK_UNDELIVERED but where the domain_dev->lldd_dev link is still
intact.  Introduce a 'gone' flag to prevent i/o while waiting for libsas
to take action on the port down event.

One 'core' leftover is that we currently call
scic_remote_device_destruct() from isci_remote_device_deconstruct()
which is called when the 'core' says the device is stopped.  It would be
more natural for the final put to trigger
isci_remote_device_deconstruct() but this implementation is deferred as
it requires other changes.
Signed-off-by: Dan Williams <dan.j.williams@intel.com>

drivers/scsi/isci/host.c

@@ -1327,7 +1327,7 @@ void isci_host_deinit(struct isci_host *ihost)
 		struct isci_remote_device *idev, *d;
 
 		list_for_each_entry_safe(idev, d, &iport->remote_dev_list, node) {
-			isci_remote_device_change_state(idev, isci_stopping);
+			if (test_bit(IDEV_ALLOCATED, &idev->flags))
 				isci_remote_device_stop(ihost, idev);
 		}
 	}

drivers/scsi/isci/port.c

@@ -321,8 +321,7 @@ static void isci_port_link_down(struct isci_host *isci_host,
 				dev_dbg(&isci_host->pdev->dev,
 					"%s: isci_device = %p\n",
 					__func__, isci_device);
-				isci_remote_device_change_state(isci_device,
-								isci_stopping);
+				set_bit(IDEV_GONE, &isci_device->flags);
 			}
 		}
 		isci_port_change_state(isci_port, isci_stopping);

drivers/scsi/isci/remote_device.c

@@ -94,7 +94,7 @@ static void isci_remote_device_not_ready(struct isci_host *ihost,
 		"%s: isci_device = %p\n", __func__, idev);
 
 	if (reason == SCIC_REMOTE_DEVICE_NOT_READY_STOP_REQUESTED)
-		isci_remote_device_change_state(idev, isci_stopping);
+		set_bit(IDEV_GONE, &idev->flags);
 	else
 		/* device ready is actually a "not ready for io" state. */
 		isci_remote_device_change_state(idev, isci_ready);
@@ -449,8 +449,10 @@ static void scic_sds_remote_device_start_request(struct scic_sds_remote_device *
 	/* cleanup requests that failed after starting on the port */
 	if (status != SCI_SUCCESS)
 		scic_sds_port_complete_io(sci_port, sci_dev, sci_req);
-	else
+	else {
+		kref_get(&sci_dev_to_idev(sci_dev)->kref);
 		scic_sds_remote_device_increment_request_count(sci_dev);
+	}
 }
 
 enum sci_status scic_sds_remote_device_start_io(struct scic_sds_controller *scic,
@@ -656,6 +658,8 @@ enum sci_status scic_sds_remote_device_complete_io(struct scic_sds_controller *s
 			"%s: Port:0x%p Device:0x%p Request:0x%p Status:0x%x "
 			"could not complete\n", __func__, sci_port,
 			sci_dev, sci_req, status);
+	else
+		isci_put_device(sci_dev_to_idev(sci_dev));
 
 	return status;
 }
@@ -860,23 +864,11 @@ static void isci_remote_device_deconstruct(struct isci_host *ihost, struct isci_
 	 * here should go through isci_remote_device_nuke_requests.
 	 * If we hit this condition, we will need a way to complete
 	 * io requests in process */
-	while (!list_empty(&idev->reqs_in_process)) {
-
-		dev_err(&ihost->pdev->dev,
-			"%s: ** request list not empty! **\n", __func__);
-		BUG();
-	}
+	BUG_ON(!list_empty(&idev->reqs_in_process));
 
 	scic_remote_device_destruct(&idev->sci);
-	idev->domain_dev->lldd_dev = NULL;
-	idev->domain_dev = NULL;
-	idev->isci_port = NULL;
 	list_del_init(&idev->node);
-
-	clear_bit(IDEV_START_PENDING, &idev->flags);
-	clear_bit(IDEV_STOP_PENDING, &idev->flags);
-	clear_bit(IDEV_EH, &idev->flags);
-	wake_up(&ihost->eventq);
+	isci_put_device(idev);
 }
 
 /**
@@ -1314,6 +1306,22 @@ isci_remote_device_alloc(struct isci_host *ihost, struct isci_port *iport)
 	return idev;
 }
 
+void isci_remote_device_release(struct kref *kref)
+{
+	struct isci_remote_device *idev = container_of(kref, typeof(*idev), kref);
+	struct isci_host *ihost = idev->isci_port->isci_host;
+
+	idev->domain_dev = NULL;
+	idev->isci_port = NULL;
+	clear_bit(IDEV_START_PENDING, &idev->flags);
+	clear_bit(IDEV_STOP_PENDING, &idev->flags);
+	clear_bit(IDEV_GONE, &idev->flags);
+	clear_bit(IDEV_EH, &idev->flags);
+	smp_mb__before_clear_bit();
+	clear_bit(IDEV_ALLOCATED, &idev->flags);
+	wake_up(&ihost->eventq);
+}
+
 /**
  * isci_remote_device_stop() - This function is called internally to stop the
  *    remote device.
@@ -1330,7 +1338,11 @@ enum sci_status isci_remote_device_stop(struct isci_host *ihost, struct isci_rem
 	dev_dbg(&ihost->pdev->dev,
 		"%s: isci_device = %p\n", __func__, idev);
 
+	spin_lock_irqsave(&ihost->scic_lock, flags);
+	idev->domain_dev->lldd_dev = NULL; /* disable new lookups */
+	set_bit(IDEV_GONE, &idev->flags);
 	isci_remote_device_change_state(idev, isci_stopping);
+	spin_unlock_irqrestore(&ihost->scic_lock, flags);
 
 	/* Kill all outstanding requests. */
 	isci_remote_device_nuke_requests(ihost, idev);
@@ -1342,14 +1354,10 @@ enum sci_status isci_remote_device_stop(struct isci_host *ihost, struct isci_rem
 	spin_unlock_irqrestore(&ihost->scic_lock, flags);
 
 	/* Wait for the stop complete callback. */
-	if (status == SCI_SUCCESS) {
+	if (WARN_ONCE(status != SCI_SUCCESS, "failed to stop device\n"))
+		/* nothing to wait for */;
+	else
 		wait_for_device_stop(ihost, idev);
-		clear_bit(IDEV_ALLOCATED, &idev->flags);
-	}
-
-	dev_dbg(&ihost->pdev->dev,
-		"%s: idev = %p - after completion wait\n",
-		__func__, idev);
 
 	return status;
 }
@@ -1416,39 +1424,33 @@ int isci_remote_device_found(struct domain_device *domain_dev)
 	if (!isci_device)
 		return -ENODEV;
 
+	kref_init(&isci_device->kref);
 	INIT_LIST_HEAD(&isci_device->node);
-	domain_dev->lldd_dev = isci_device;
+
+	spin_lock_irq(&isci_host->scic_lock);
 	isci_device->domain_dev = domain_dev;
 	isci_device->isci_port = isci_port;
 	isci_remote_device_change_state(isci_device, isci_starting);
-
-
-	spin_lock_irq(&isci_host->scic_lock);
 	list_add_tail(&isci_device->node, &isci_port->remote_dev_list);
 
 	set_bit(IDEV_START_PENDING, &isci_device->flags);
 	status = isci_remote_device_construct(isci_port, isci_device);
-	spin_unlock_irq(&isci_host->scic_lock);
 
 	dev_dbg(&isci_host->pdev->dev,
 		"%s: isci_device = %p\n",
 		__func__, isci_device);
 
-	if (status != SCI_SUCCESS) {
-
-		spin_lock_irq(&isci_host->scic_lock);
-		isci_remote_device_deconstruct(
-			isci_host,
-			isci_device
-			);
+	if (status == SCI_SUCCESS) {
+		/* device came up, advertise it to the world */
+		domain_dev->lldd_dev = isci_device;
+	} else
+		isci_put_device(isci_device);
 	spin_unlock_irq(&isci_host->scic_lock);
-		return -ENODEV;
-	}
 
 	/* wait for the device ready callback. */
 	wait_for_device_start(isci_host, isci_device);
 
-	return 0;
+	return status == SCI_SUCCESS ? 0 : -ENODEV;
 }
 /**
  * isci_device_is_reset_pending() - This function will check if there is any

drivers/scsi/isci/remote_device.h

@@ -56,6 +56,7 @@
 #ifndef _ISCI_REMOTE_DEVICE_H_
 #define _ISCI_REMOTE_DEVICE_H_
 #include <scsi/libsas.h>
+#include <linux/kref.h>
 #include "scu_remote_node_context.h"
 #include "remote_node_context.h"
 #include "port.h"
@@ -134,7 +135,9 @@ struct isci_remote_device {
 	#define IDEV_STOP_PENDING 1
 	#define IDEV_ALLOCATED 2
 	#define IDEV_EH 3
+	#define IDEV_GONE 4
 	unsigned long flags;
+	struct kref kref;
 	struct isci_port *isci_port;
 	struct domain_device *domain_dev;
 	struct list_head node;
@@ -145,6 +148,26 @@ struct isci_remote_device {
 
 #define ISCI_REMOTE_DEVICE_START_TIMEOUT 5000
 
+/* device reference routines must be called under scic_lock */
+static inline struct isci_remote_device *isci_lookup_device(struct domain_device *dev)
+{
+	struct isci_remote_device *idev = dev->lldd_dev;
+
+	if (idev && !test_bit(IDEV_GONE, &idev->flags)) {
+		kref_get(&idev->kref);
+		return idev;
+	}
+
+	return NULL;
+}
+
+void isci_remote_device_release(struct kref *kref);
+static inline void isci_put_device(struct isci_remote_device *idev)
+{
+	if (idev)
+		kref_put(&idev->kref, isci_remote_device_release);
+}
+
 enum sci_status isci_remote_device_stop(struct isci_host *ihost,
 					struct isci_remote_device *idev);
 void isci_remote_device_nuke_requests(struct isci_host *ihost,

drivers/scsi/isci/request.c

@@ -2313,7 +2313,7 @@ static void isci_request_set_open_reject_status(
  * none.
  */
 static void isci_request_handle_controller_specific_errors(
-	struct isci_remote_device *isci_device,
+	struct isci_remote_device *idev,
 	struct isci_request *request,
 	struct sas_task *task,
 	enum service_response *response_ptr,
@@ -2353,8 +2353,7 @@ static void isci_request_handle_controller_specific_errors(
 			 * that we ignore the quiesce state, since we are
 			 * concerned about the actual device state.
 			 */
-			if ((isci_device->status == isci_stopping) ||
-			    (isci_device->status == isci_stopped))
+			if (!idev)
 				*status_ptr = SAS_DEVICE_UNKNOWN;
 			else
 				*status_ptr = SAS_ABORTED_TASK;
@@ -2367,8 +2366,7 @@ static void isci_request_handle_controller_specific_errors(
 			/* Task in the target is not done. */
 			*response_ptr = SAS_TASK_UNDELIVERED;
 
-			if ((isci_device->status == isci_stopping) ||
-			    (isci_device->status == isci_stopped))
+			if (!idev)
 				*status_ptr = SAS_DEVICE_UNKNOWN;
 			else
 				*status_ptr = SAM_STAT_TASK_ABORTED;
@@ -2399,8 +2397,7 @@ static void isci_request_handle_controller_specific_errors(
 		 * that we ignore the quiesce state, since we are
 		 * concerned about the actual device state.
 		 */
-		if ((isci_device->status == isci_stopping) ||
-		    (isci_device->status == isci_stopped))
+		if (!idev)
 			*status_ptr = SAS_DEVICE_UNKNOWN;
 		else
 			*status_ptr = SAS_ABORTED_TASK;
@@ -2629,7 +2626,7 @@ static void isci_request_io_request_complete(struct isci_host *isci_host,
 	struct ssp_response_iu *resp_iu;
 	void *resp_buf;
 	unsigned long task_flags;
-	struct isci_remote_device *isci_device   = request->isci_device;
+	struct isci_remote_device *idev = isci_lookup_device(task->dev);
 	enum service_response response       = SAS_TASK_UNDELIVERED;
 	enum exec_status status         = SAS_ABORTED_TASK;
 	enum isci_request_status request_status;
@@ -2672,9 +2669,7 @@ static void isci_request_io_request_complete(struct isci_host *isci_host,
 		 * that we ignore the quiesce state, since we are
 		 * concerned about the actual device state.
 		 */
-		if ((isci_device->status == isci_stopping)
-		    || (isci_device->status == isci_stopped)
-		    )
+		if (!idev)
 			status = SAS_DEVICE_UNKNOWN;
 		else
 			status = SAS_ABORTED_TASK;
@@ -2697,8 +2692,7 @@ static void isci_request_io_request_complete(struct isci_host *isci_host,
 		request->complete_in_target = true;
 		response = SAS_TASK_UNDELIVERED;
 
-		if ((isci_device->status == isci_stopping) ||
-		    (isci_device->status == isci_stopped))
+		if (!idev)
 			/* The device has been /is being stopped. Note that
 			 * we ignore the quiesce state, since we are
 			 * concerned about the actual device state.
@@ -2728,8 +2722,7 @@ static void isci_request_io_request_complete(struct isci_host *isci_host,
 		 * that we ignore the quiesce state, since we are
 		 * concerned about the actual device state.
 		 */
-		if ((isci_device->status == isci_stopping) ||
-		    (isci_device->status == isci_stopped))
+		if (!idev)
 			status = SAS_DEVICE_UNKNOWN;
 		else
 			status = SAS_ABORTED_TASK;
@@ -2861,8 +2854,7 @@ static void isci_request_io_request_complete(struct isci_host *isci_host,
 			 * that we ignore the quiesce state, since we are
 			 * concerned about the actual device state.
 			 */
-			if ((isci_device->status == isci_stopping) ||
-			    (isci_device->status == isci_stopped))
+			if (!idev)
 				status = SAS_DEVICE_UNKNOWN;
 			else
 				status = SAS_ABORTED_TASK;
@@ -2873,7 +2865,7 @@ static void isci_request_io_request_complete(struct isci_host *isci_host,
 		case SCI_FAILURE_CONTROLLER_SPECIFIC_IO_ERR:
 
 			isci_request_handle_controller_specific_errors(
-				isci_device, request, task, &response, &status,
+				idev, request, task, &response, &status,
 				&complete_to_host);
 
 			break;
@@ -2902,8 +2894,7 @@ static void isci_request_io_request_complete(struct isci_host *isci_host,
 
 			/* Fail the I/O so it can be retried. */
 			response = SAS_TASK_UNDELIVERED;
-			if ((isci_device->status == isci_stopping) ||
-			    (isci_device->status == isci_stopped))
+			if (!idev)
 				status = SAS_DEVICE_UNKNOWN;
 			else
 				status = SAS_ABORTED_TASK;
@@ -2926,8 +2917,7 @@ static void isci_request_io_request_complete(struct isci_host *isci_host,
 			 * that we ignore the quiesce state, since we are
 			 * concerned about the actual device state.
 			 */
-			if ((isci_device->status == isci_stopping) ||
-			    (isci_device->status == isci_stopped))
+			if (!idev)
 				status = SAS_DEVICE_UNKNOWN;
 			else
 				status = SAS_ABORTED_TASK;
@@ -2953,8 +2943,10 @@ static void isci_request_io_request_complete(struct isci_host *isci_host,
 
 	/* complete the io request to the core. */
 	scic_controller_complete_io(&isci_host->sci,
-				    &isci_device->sci,
+				    request->sci.target_device,
 				    &request->sci);
+	isci_put_device(idev);
+
 	/* set terminated handle so it cannot be completed or
 	 * terminated again, and to cause any calls into abort
 	 * task to recognize the already completed case.
@@ -3511,7 +3503,6 @@ static enum sci_status isci_io_request_build(
 }
 
 static struct isci_request *isci_request_alloc_core(struct isci_host *ihost,
-						    struct isci_remote_device *idev,
 						    gfp_t gfp_flags)
 {
 	dma_addr_t handle;
@@ -3528,7 +3519,6 @@ static struct isci_request *isci_request_alloc_core(struct isci_host *ihost,
 	spin_lock_init(&ireq->state_lock);
 	ireq->request_daddr = handle;
 	ireq->isci_host = ihost;
-	ireq->isci_device = idev;
 	ireq->io_request_completion = NULL;
 	ireq->terminated = false;
 
@@ -3546,12 +3536,11 @@ static struct isci_request *isci_request_alloc_core(struct isci_host *ihost,
 
 static struct isci_request *isci_request_alloc_io(struct isci_host *ihost,
 						  struct sas_task *task,
-						  struct isci_remote_device *idev,
 						  gfp_t gfp_flags)
 {
 	struct isci_request *ireq;
 
-	ireq = isci_request_alloc_core(ihost, idev, gfp_flags);
+	ireq = isci_request_alloc_core(ihost, gfp_flags);
 	if (ireq) {
 		ireq->ttype_ptr.io_task_ptr = task;
 		ireq->ttype = io_task;
@@ -3562,12 +3551,11 @@ static struct isci_request *isci_request_alloc_io(struct isci_host *ihost,
 
 struct isci_request *isci_request_alloc_tmf(struct isci_host *ihost,
 					    struct isci_tmf *isci_tmf,
-					    struct isci_remote_device *idev,
 					    gfp_t gfp_flags)
 {
 	struct isci_request *ireq;
 
-	ireq = isci_request_alloc_core(ihost, idev, gfp_flags);
+	ireq = isci_request_alloc_core(ihost, gfp_flags);
 	if (ireq) {
 		ireq->ttype_ptr.tmf_task_ptr = isci_tmf;
 		ireq->ttype = tmf_task;
@@ -3575,21 +3563,16 @@ struct isci_request *isci_request_alloc_tmf(struct isci_host *ihost,
 	return ireq;
 }
 
-int isci_request_execute(struct isci_host *ihost, struct sas_task *task,
-			 gfp_t gfp_flags)
+int isci_request_execute(struct isci_host *ihost, struct isci_remote_device *idev,
+			 struct sas_task *task, gfp_t gfp_flags)
 {
 	enum sci_status status = SCI_FAILURE_UNSUPPORTED_PROTOCOL;
-	struct scic_sds_remote_device *sci_dev;
-	struct isci_remote_device *idev;
 	struct isci_request *ireq;
 	unsigned long flags;
 	int ret = 0;
 
-	idev = task->dev->lldd_dev;
-	sci_dev = &idev->sci;
-
 	/* do common allocation and init of request object. */
-	ireq = isci_request_alloc_io(ihost, task, idev, gfp_flags);
+	ireq = isci_request_alloc_io(ihost, task, gfp_flags);
 	if (!ireq)
 		goto out;
 
@@ -3605,8 +3588,7 @@ int isci_request_execute(struct isci_host *ihost, struct sas_task *task,
 	spin_lock_irqsave(&ihost->scic_lock, flags);
 
 	/* send the request, let the core assign the IO TAG.	*/
-	status = scic_controller_start_io(&ihost->sci, sci_dev,
-					  &ireq->sci,
+	status = scic_controller_start_io(&ihost->sci, &idev->sci, &ireq->sci,
 					  SCI_CONTROLLER_INVALID_IO_TAG);
 	if (status != SCI_SUCCESS &&
 	    status != SCI_FAILURE_REMOTE_DEVICE_RESET_REQUIRED) {

drivers/scsi/isci/request.h

@@ -285,7 +285,6 @@ struct isci_request {
 		struct isci_tmf *tmf_task_ptr;  /* When ttype==tmf_task */
 	} ttype_ptr;
 	struct isci_host *isci_host;
-	struct isci_remote_device *isci_device;
 	/* For use in the requests_to_{complete|abort} lists: */
 	struct list_head completed_node;
 	/* For use in the reqs_in_process list: */
@@ -681,12 +680,10 @@ static inline void isci_request_free(struct isci_host *isci_host,
 
 struct isci_request *isci_request_alloc_tmf(struct isci_host *ihost,
 					    struct isci_tmf *isci_tmf,
-					    struct isci_remote_device *idev,
 					    gfp_t gfp_flags);
 
-int isci_request_execute(struct isci_host *isci_host,
-			 struct sas_task *task,
-			 gfp_t gfp_flags);
+int isci_request_execute(struct isci_host *ihost, struct isci_remote_device *idev,
+			 struct sas_task *task, gfp_t gfp_flags);
 
 /**
  * isci_request_unmap_sgl() - This function unmaps the DMA address of a given

drivers/scsi/isci/sata.c

@@ -213,11 +213,10 @@ int isci_task_send_lu_reset_sata(
 
 	/* Send the soft reset to the target */
 	#define ISCI_SRST_TIMEOUT_MS 25000 /* 25 second timeout. */
-	isci_task_build_tmf(&tmf, isci_device, isci_tmf_sata_srst_high,
-			    NULL, NULL
-			    );
+	isci_task_build_tmf(&tmf, isci_tmf_sata_srst_high, NULL, NULL);
 
-	ret = isci_task_execute_tmf(isci_host, &tmf, ISCI_SRST_TIMEOUT_MS);
+	ret = isci_task_execute_tmf(isci_host, isci_device, &tmf,
+				    ISCI_SRST_TIMEOUT_MS);
 
 	if (ret != TMF_RESP_FUNC_COMPLETE) {
 		dev_warn(&isci_host->pdev->dev,

drivers/scsi/isci/task.h

@@ -213,18 +213,15 @@ int isci_bus_reset_handler(struct scsi_cmnd *cmd);
 
 void isci_task_build_tmf(
 	struct isci_tmf *tmf,
-	struct isci_remote_device *isci_device,
 	enum isci_tmf_function_codes code,
 	void (*tmf_sent_cb)(enum isci_tmf_cb_state,
 			    struct isci_tmf *,
 			    void *),
 	void *cb_data);
 
-
-int isci_task_execute_tmf(
-	struct isci_host *isci_host,
-	struct isci_tmf *tmf,
-	unsigned long timeout_ms);
+int isci_task_execute_tmf(struct isci_host *isci_host,
+			  struct isci_remote_device *idev,
+			  struct isci_tmf *tmf, unsigned long timeout_ms);
 
 /**
  * enum isci_completion_selection - This enum defines the possible actions to