Commit 23bace67 authored by Ofir Bitton's avatar Ofir Bitton Committed by Oded Gabbay
Browse files

habanalabs: allow reset upon device release 



We introduce a new type of reset which is reset upon device release.
This reset is very similar to soft reset except the fact it is
performed only upon device release and not upon user sysfs request
nor TDR.

The purpose of this reset is to make sure the device is returned to
IDLE state after the current user has finished working with the device.

Signed-off-by: default avatarOfir Bitton <obitton@habana.ai>
Reviewed-by: default avatarOded Gabbay <ogabbay@kernel.org>
Signed-off-by: default avatarOded Gabbay <ogabbay@kernel.org>
parent 4d041216

drivers/misc/habanalabs/common/device.c

+21 −4
Original line number Diff line number Diff line
@@ -86,7 +86,7 @@ static void hpriv_release(struct kref *ref) 


	if ((hdev->reset_if_device_not_idle && !device_is_idle)

			|| hdev->reset_upon_device_release)

		hl_device_reset(hdev, 0);

		hl_device_reset(hdev, HL_RESET_DEVICE_RELEASE);

}



void hl_hpriv_get(struct hl_fpriv *hpriv)
@@ -885,7 +885,7 @@ static void device_disable_open_processes(struct hl_device *hdev) 
int hl_device_reset(struct hl_device *hdev, u32 flags)

{

	u64 idle_mask[HL_BUSY_ENGINES_MASK_EXT_SIZE] = {0};

	bool hard_reset, from_hard_reset_thread;

	bool hard_reset, from_hard_reset_thread, hard_instead_soft = false;

	int i, rc;



	if (!hdev->init_done) {
@@ -897,11 +897,28 @@ int hl_device_reset(struct hl_device *hdev, u32 flags) 
	hard_reset = (flags & HL_RESET_HARD) != 0;

	from_hard_reset_thread = (flags & HL_RESET_FROM_RESET_THREAD) != 0;



	if ((!hard_reset) && (!hdev->supports_soft_reset)) {

		dev_dbg(hdev->dev, "Doing hard-reset instead of soft-reset\n");

	if (!hard_reset && !hdev->supports_soft_reset) {

		hard_instead_soft = true;

		hard_reset = true;

	}



	if (hdev->reset_upon_device_release &&

			(flags & HL_RESET_DEVICE_RELEASE)) {

		dev_dbg(hdev->dev,

			"Perform %s-reset upon device release\n",

			hard_reset ? "hard" : "soft");

		goto do_reset;

	}



	if (!hard_reset && !hdev->allow_external_soft_reset) {

		hard_instead_soft = true;

		hard_reset = true;

	}



	if (hard_instead_soft)

		dev_dbg(hdev->dev, "Doing hard-reset instead of soft-reset\n");



do_reset:

	/* Re-entry of reset thread */

	if (from_hard_reset_thread && hdev->process_kill_trial_cnt)

		goto kill_processes;

drivers/misc/habanalabs/common/habanalabs.h

+7 −0
Original line number Diff line number Diff line
@@ -119,11 +119,15 @@ enum hl_mmu_page_table_location { 
 *

 * - HL_RESET_TDR

 *       Set if reset is due to TDR

 *

 * - HL_RESET_DEVICE_RELEASE

 *       Set if reset is due to device release

 */

#define HL_RESET_HARD			(1 << 0)

#define HL_RESET_FROM_RESET_THREAD	(1 << 1)

#define HL_RESET_HEARTBEAT		(1 << 2)

#define HL_RESET_TDR			(1 << 3)

#define HL_RESET_DEVICE_RELEASE		(1 << 4)



#define HL_MAX_SOBS_PER_MONITOR	8
@@ -2181,6 +2185,8 @@ struct hl_mmu_funcs { 
 * @collective_mon_idx: helper index for collective initialization

 * @supports_coresight: is CoreSight supported.

 * @supports_soft_reset: is soft reset supported.

 * @allow_external_soft_reset: true if soft reset initiated by user or TDR is

 *                             allowed.

 * @supports_cb_mapping: is mapping a CB to the device's MMU supported.

 * @needs_reset: true if reset_on_lockup is false and device should be reset

 *               due to lockup.
@@ -2301,6 +2307,7 @@ struct hl_device { 
	u8				collective_mon_idx;

	u8				supports_coresight;

	u8				supports_soft_reset;

	u8				allow_external_soft_reset;

	u8				supports_cb_mapping;

	u8				needs_reset;

	u8				process_kill_trial_cnt;

drivers/misc/habanalabs/common/sysfs.c

+1 −1
Original line number Diff line number Diff line
@@ -208,7 +208,7 @@ static ssize_t soft_reset_store(struct device *dev, 
		goto out;

	}



	if (!hdev->supports_soft_reset) {

	if (!hdev->allow_external_soft_reset) {

		dev_err(hdev->dev, "Device does not support soft-reset\n");

		goto out;

	}

drivers/misc/habanalabs/goya/goya.c

+1 −0
Original line number Diff line number Diff line
@@ -954,6 +954,7 @@ static int goya_sw_init(struct hl_device *hdev) 
	spin_lock_init(&goya->hw_queues_lock);

	hdev->supports_coresight = true;

	hdev->supports_soft_reset = true;

	hdev->allow_external_soft_reset = true;



	goya_set_pci_memory_regions(hdev);