drm/amdkfd: Avoid using doorbell_off as offset in process doorbell pages

dorbell_off in the queue properties is mainly used for the doorbell dw
offset in pci bar. We should not set it to the doorbell byte offset in
process doorbell pages. This makes the code much easier to read.

Signed-off-by: Yong Zhao <Yong.Zhao@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

4 files changed, 16 insertions, 11 deletions

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
index 568989b7947f..1041f4d627a6 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
@@ -258,6 +258,7 @@ static int kfd_ioctl_create_queue(struct file *filep, struct kfd_process *p,
 	unsigned int queue_id;
 	struct kfd_process_device *pdd;
 	struct queue_properties q_properties;
+	uint32_t doorbell_offset_in_process = 0;
 
 	memset(&q_properties, 0, sizeof(struct queue_properties));
 
@@ -286,7 +287,8 @@ static int kfd_ioctl_create_queue(struct file *filep, struct kfd_process *p,
 			p->pasid,
 			dev->id);
 
-	err = pqm_create_queue(&p->pqm, dev, filep, &q_properties, &queue_id);
+	err = pqm_create_queue(&p->pqm, dev, filep, &q_properties, &queue_id,
+			&doorbell_offset_in_process);
 	if (err != 0)
 		goto err_create_queue;
 
@@ -297,12 +299,10 @@ static int kfd_ioctl_create_queue(struct file *filep, struct kfd_process *p,
 	args->doorbell_offset = KFD_MMAP_TYPE_DOORBELL;
 	args->doorbell_offset |= KFD_MMAP_GPU_ID(args->gpu_id);
 	if (KFD_IS_SOC15(dev->device_info->asic_family))
-		/* On SOC15 ASICs, doorbell allocation must be
-		 * per-device, and independent from the per-process
-		 * queue_id. Return the doorbell offset within the
-		 * doorbell aperture to user mode.
+		/* On SOC15 ASICs, include the doorbell offset within the
+		 * process doorbell frame, which is 2 pages.
 		 */
-		args->doorbell_offset |= q_properties.doorbell_off;
+		args->doorbell_offset |= doorbell_offset_in_process;
 
 	mutex_unlock(&p->mutex);
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.c b/drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.c
index d59f2cd056c6..1d33c4f25263 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_dbgdev.c
@@ -185,7 +185,7 @@ static int dbgdev_register_diq(struct kfd_dbgdev *dbgdev)
 	properties.type = KFD_QUEUE_TYPE_DIQ;
 
 	status = pqm_create_queue(dbgdev->pqm, dbgdev->dev, NULL,
-				&properties, &qid);
+				&properties, &qid, NULL);
 
 	if (status) {
 		pr_err("Failed to create DIQ\n");
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
index 603d793a8f56..8ed691fb345a 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
@@ -902,7 +902,8 @@ int pqm_create_queue(struct process_queue_manager *pqm,
 			    struct kfd_dev *dev,
 			    struct file *f,
 			    struct queue_properties *properties,
-			    unsigned int *qid);
+			    unsigned int *qid,
+			    uint32_t *p_doorbell_offset_in_process);
 int pqm_destroy_queue(struct process_queue_manager *pqm, unsigned int qid);
 int pqm_update_queue(struct process_queue_manager *pqm, unsigned int qid,
 			struct queue_properties *p);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c
index 2659d226c056..8bc69689f9b8 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c
@@ -192,7 +192,8 @@ int pqm_create_queue(struct process_queue_manager *pqm,
 			    struct kfd_dev *dev,
 			    struct file *f,
 			    struct queue_properties *properties,
-			    unsigned int *qid)
+			    unsigned int *qid,
+			    uint32_t *p_doorbell_offset_in_process)
 {
 	int retval;
 	struct kfd_process_device *pdd;
@@ -303,12 +304,15 @@ int pqm_create_queue(struct process_queue_manager *pqm,
 		goto err_create_queue;
 	}
 
-	if (q)
+	if (q && p_doorbell_offset_in_process)
 		/* Return the doorbell offset within the doorbell page
 		 * to the caller so it can be passed up to user mode
 		 * (in bytes).
+		 * There are always 1024 doorbells per process, so in case
+		 * of 8-byte doorbells, there are two doorbell pages per
+		 * process.
 		 */
-		properties->doorbell_off =
+		*p_doorbell_offset_in_process =
 			(q->properties.doorbell_off * sizeof(uint32_t)) &
 			(kfd_doorbell_process_slice(dev) - 1);