drm/amdkfd: Adjust dequeue latencies and timeouts

Adjust latencies and timeouts for dequeueing with HWS and consolidate
them in one place. Make them longer to allow long running waves to
complete without causing a timeout. The timeout is twice as long as the
latency plus some buffer to make sure we don't detect a timeout
prematurely.

Change timeouts for dequeueing HQDs without HWS. KFD_UNMAP_LATENCY is
more consistent with what the HWS does for user queues.

Signed-off-by: Yong Zhao <yong.zhao@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>

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

@@ -323,7 +323,7 @@ static int destroy_queue_nocpsch(struct device_queue_manager *dqm,
 
 	retval = mqd->destroy_mqd(mqd, q->mqd,
 				KFD_PREEMPT_TYPE_WAVEFRONT_RESET,
-				QUEUE_PREEMPT_DEFAULT_TIMEOUT_MS,
+				KFD_UNMAP_LATENCY_MS,
 				q->pipe, q->queue);
 
 	if (retval)

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

@@ -29,7 +29,9 @@
 #include "kfd_priv.h"
 #include "kfd_mqd_manager.h"
 
-#define QUEUE_PREEMPT_DEFAULT_TIMEOUT_MS	(500)
+#define KFD_UNMAP_LATENCY_MS			(4000)
+#define QUEUE_PREEMPT_DEFAULT_TIMEOUT_MS (2 * KFD_UNMAP_LATENCY_MS + 1000)
+
 #define CIK_VMID_NUM				(8)
 #define KFD_VMID_START_OFFSET			(8)
 #define VMID_PER_DEVICE				CIK_VMID_NUM

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

@@ -185,7 +185,7 @@ static void uninitialize(struct kernel_queue *kq)
 		kq->mqd->destroy_mqd(kq->mqd,
 					kq->queue->mqd,
 					KFD_PREEMPT_TYPE_WAVEFRONT_RESET,
-					QUEUE_PREEMPT_DEFAULT_TIMEOUT_MS,
+					KFD_UNMAP_LATENCY_MS,
 					kq->queue->pipe,
 					kq->queue->queue);
 	else if (kq->queue->properties.type == KFD_QUEUE_TYPE_DIQ)

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

@@ -376,7 +376,7 @@ int pm_send_set_resources(struct packet_manager *pm,
 	packet->bitfields2.queue_type =
 			queue_type__mes_set_resources__hsa_interface_queue_hiq;
 	packet->bitfields2.vmid_mask = res->vmid_mask;
-	packet->bitfields2.unmap_latency = KFD_UNMAP_LATENCY;
+	packet->bitfields2.unmap_latency = KFD_UNMAP_LATENCY_MS / 100;
 	packet->bitfields7.oac_mask = res->oac_mask;
 	packet->bitfields8.gds_heap_base = res->gds_heap_base;
 	packet->bitfields8.gds_heap_size = res->gds_heap_size;

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

@@ -676,11 +676,8 @@ int amdkfd_fence_wait_timeout(unsigned int *fence_addr,
 
 /* Packet Manager */
 
-#define KFD_HIQ_TIMEOUT (500)
-
 #define KFD_FENCE_COMPLETED (100)
 #define KFD_FENCE_INIT   (10)
-#define KFD_UNMAP_LATENCY (150)
 
 struct packet_manager {
 	struct device_queue_manager *dqm;