Commit 51a0f459 authored by Oak Zeng's avatar Oak Zeng Committed by Alex Deucher
Browse files

drm/amdkfd: Check HIQ's MQD for queue preemption status 



MEC firmware can silently fail the queue preemption request
without time out. In this case, HIQ's MQD's queue_doorbell_id
will be set. Check this field to see whether last queue preemption
was successful or not.

Signed-off-by: default avatarOak Zeng <Oak.Zeng@amd.com>
Suggested-by: default avatarJay Cornwall <Jay.Cornwall@amd.com>
Acked-by: default avatarFelix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: default avatarAlex Deucher <alexander.deucher@amd.com>
parent 6d909c5d

drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c

+17 −0
Original line number Diff line number Diff line
@@ -1393,6 +1393,7 @@ static int unmap_queues_cpsch(struct device_queue_manager *dqm, 
				uint32_t filter_param)

{

	int retval = 0;

	struct mqd_manager *mqd_mgr;



	if (!dqm->sched_running)

		return 0;
@@ -1424,6 +1425,22 @@ static int unmap_queues_cpsch(struct device_queue_manager *dqm, 
		return retval;

	}



	/* In the current MEC firmware implementation, if compute queue

	 * doesn't response to the preemption request in time, HIQ will

	 * abandon the unmap request without returning any timeout error

	 * to driver. Instead, MEC firmware will log the doorbell of the

	 * unresponding compute queue to HIQ.MQD.queue_doorbell_id fields.

	 * To make sure the queue unmap was successful, driver need to

	 * check those fields

	 */

	mqd_mgr = dqm->mqd_mgrs[KFD_MQD_TYPE_HIQ];

	if (mqd_mgr->read_doorbell_id(dqm->packets.priv_queue->queue->mqd)) {

		pr_err("HIQ MQD's queue_doorbell_id0 is not 0, Queue preemption time out\n");

		while (halt_if_hws_hang)

			schedule();

		return -ETIME;

	}



	pm_release_ib(&dqm->packets);

	dqm->active_runlist = false;

drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.h

+1 −0
Original line number Diff line number Diff line
@@ -101,6 +101,7 @@ struct mqd_manager { 
#if defined(CONFIG_DEBUG_FS)

	int	(*debugfs_show_mqd)(struct seq_file *m, void *data);

#endif

	uint32_t (*read_doorbell_id)(void *mqd);



	struct mutex	mqd_mutex;

	struct kfd_dev	*dev;

drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c

+8 −0
Original line number Diff line number Diff line
@@ -226,6 +226,13 @@ static void update_mqd(struct mqd_manager *mm, void *mqd, 
	__update_mqd(mm, mqd, q, 1);

}



static uint32_t read_doorbell_id(void *mqd)

{

	struct cik_mqd *m = (struct cik_mqd *)mqd;



	return m->queue_doorbell_id0;

}



static void update_mqd_hawaii(struct mqd_manager *mm, void *mqd,

			struct queue_properties *q)

{
@@ -398,6 +405,7 @@ struct mqd_manager *mqd_manager_init_cik(enum KFD_MQD_TYPE type, 
#if defined(CONFIG_DEBUG_FS)

		mqd->debugfs_show_mqd = debugfs_show_mqd;

#endif

		mqd->read_doorbell_id = read_doorbell_id;

		break;

	case KFD_MQD_TYPE_DIQ:

		mqd->allocate_mqd = allocate_mqd;

drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v10.c

+8 −0
Original line number Diff line number Diff line
@@ -224,6 +224,13 @@ static void update_mqd(struct mqd_manager *mm, void *mqd, 
	q->is_active = QUEUE_IS_ACTIVE(*q);

}



static uint32_t read_doorbell_id(void *mqd)

{

	struct v10_compute_mqd *m = (struct v10_compute_mqd *)mqd;



	return m->queue_doorbell_id0;

}



static int destroy_mqd(struct mqd_manager *mm, void *mqd,

		       enum kfd_preempt_type type,

		       unsigned int timeout, uint32_t pipe_id,
@@ -425,6 +432,7 @@ struct mqd_manager *mqd_manager_init_v10(enum KFD_MQD_TYPE type, 
#if defined(CONFIG_DEBUG_FS)

		mqd->debugfs_show_mqd = debugfs_show_mqd;

#endif

		mqd->read_doorbell_id = read_doorbell_id;

		pr_debug("%s@%i\n", __func__, __LINE__);

		break;

	case KFD_MQD_TYPE_DIQ:

drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c

+8 −0
Original line number Diff line number Diff line
@@ -276,6 +276,13 @@ static void update_mqd(struct mqd_manager *mm, void *mqd, 
}





static uint32_t read_doorbell_id(void *mqd)

{

	struct v9_mqd *m = (struct v9_mqd *)mqd;



	return m->queue_doorbell_id0;

}



static int destroy_mqd(struct mqd_manager *mm, void *mqd,

			enum kfd_preempt_type type,

			unsigned int timeout, uint32_t pipe_id,
@@ -477,6 +484,7 @@ struct mqd_manager *mqd_manager_init_v9(enum KFD_MQD_TYPE type, 
#if defined(CONFIG_DEBUG_FS)

		mqd->debugfs_show_mqd = debugfs_show_mqd;

#endif

		mqd->read_doorbell_id = read_doorbell_id;

		break;

	case KFD_MQD_TYPE_DIQ:

		mqd->allocate_mqd = allocate_mqd;