Commit 1eebb259 authored by Tomer Tayar's avatar Tomer Tayar Committed by Oded Gabbay
Browse files

habanalabs: move reset workqueue to be under hl_device 



'struct hl_device_reset_work' is used as a wrapper for the reset work
and its parameters, including the reset workqueue on which it runs.
In a future commit, another reset related work with similar parameters
is going to be added, but it won't use the reset workqueue.

As in any case there is a single reset workqueue, and to allow the resue
of this structure, move the reset workqueue to 'struct hl_device'.

Signed-off-by: default avatarTomer Tayar <ttayar@habana.ai>
Reviewed-by: default avatarOded Gabbay <ogabbay@kernel.org>
Signed-off-by: default avatarOded Gabbay <ogabbay@kernel.org>
parent 51236cd9

drivers/misc/habanalabs/common/device.c

+6 −9
Original line number Diff line number Diff line
@@ -684,8 +684,7 @@ static void device_hard_reset_pending(struct work_struct *work) 
			"Could not reset device. will try again in %u seconds",

			HL_PENDING_RESET_PER_SEC);



		queue_delayed_work(device_reset_work->wq,

			&device_reset_work->reset_work,

		queue_delayed_work(hdev->reset_wq, &device_reset_work->reset_work,

					msecs_to_jiffies(HL_PENDING_RESET_PER_SEC * 1000));

	}

}
@@ -801,9 +800,8 @@ static int device_early_init(struct hl_device *hdev) 


	hl_mem_mgr_init(hdev->dev, &hdev->kernel_mem_mgr);



	hdev->device_reset_work.wq =

			create_singlethread_workqueue("hl_device_reset");

	if (!hdev->device_reset_work.wq) {

	hdev->reset_wq = create_singlethread_workqueue("hl_device_reset");

	if (!hdev->reset_wq) {

		rc = -ENOMEM;

		dev_err(hdev->dev, "Failed to create device reset WQ\n");

		goto free_cb_mgr;
@@ -879,7 +877,7 @@ static void device_early_fini(struct hl_device *hdev) 
	destroy_workqueue(hdev->ts_free_obj_wq);

	destroy_workqueue(hdev->cs_cmplt_wq);

	destroy_workqueue(hdev->eq_wq);

	destroy_workqueue(hdev->device_reset_work.wq);

	destroy_workqueue(hdev->reset_wq);



	for (i = 0 ; i < hdev->asic_prop.completion_queues_count ; i++)

		destroy_workqueue(hdev->cq_wq[i]);
@@ -1460,8 +1458,7 @@ int hl_device_reset(struct hl_device *hdev, u32 flags) 
		 * Because the reset function can't run from heartbeat work,

		 * we need to call the reset function from a dedicated work.

		 */

		queue_delayed_work(hdev->device_reset_work.wq,

			&hdev->device_reset_work.reset_work, 0);

		queue_delayed_work(hdev->reset_wq, &hdev->device_reset_work.reset_work, 0);



		return 0;

	}

drivers/misc/habanalabs/common/habanalabs.h

+6 −6
Original line number Diff line number Diff line
@@ -2682,14 +2682,12 @@ void hl_wreg(struct hl_device *hdev, u32 reg, u32 val); 
struct hwmon_chip_info;



/**

 * struct hl_device_reset_work - reset workqueue task wrapper.

 * @wq: work queue for device reset procedure.

 * struct hl_device_reset_work - reset work wrapper.

 * @reset_work: reset work to be done.

 * @hdev: habanalabs device structure.

 * @flags: reset flags.

 */

struct hl_device_reset_work {

	struct workqueue_struct		*wq;

	struct delayed_work	reset_work;

	struct hl_device	*hdev;

	u32			flags;
@@ -3061,6 +3059,7 @@ struct hl_reset_info { 
 *               context.

 * @ts_free_obj_wq: work queue for timestamp registration objects release.

 * @prefetch_wq: work queue for MMU pre-fetch operations.

 * @reset_wq: work queue for device reset procedure.

 * @kernel_ctx: Kernel driver context structure.

 * @kernel_queues: array of hl_hw_queue.

 * @cs_mirror_list: CS mirror list for TDR.
@@ -3232,6 +3231,7 @@ struct hl_device { 
	struct workqueue_struct		*cs_cmplt_wq;

	struct workqueue_struct		*ts_free_obj_wq;

	struct workqueue_struct		*prefetch_wq;

	struct workqueue_struct		*reset_wq;

	struct hl_ctx			*kernel_ctx;

	struct hl_hw_queue		*kernel_queues;

	struct list_head		cs_mirror_list;