drm/amdkfd: update SIMD distribution algo for GFXIP 9.4.2 onwards

In certain cooperative group dispatch scenarios the default SPI resource
allocation may cause reduced per-CU workgroup occupancy. Set
COMPUTE_RESOURCE_LIMITS.FORCE_SIMD_DIST=1 to mitigate soft hang
scenarions.
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Suggested-by: Joseph Greathouse <Joseph.Greathouse@amd.com>
Signed-off-by: Rajneesh Bhardwaj <rajneesh.bhardwaj@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

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

@@ -303,6 +303,15 @@ static void update_mqd(struct mqd_manager *mm, void *mqd,
 		update_cu_mask(mm, mqd, minfo, 0);
 	set_priority(m, q);
 
+	if (minfo && KFD_GC_VERSION(mm->dev) >= IP_VERSION(9, 4, 2)) {
+		if (minfo->update_flag & UPDATE_FLAG_IS_GWS)
+			m->compute_resource_limits |=
+				COMPUTE_RESOURCE_LIMITS__FORCE_SIMD_DIST_MASK;
+		else
+			m->compute_resource_limits &=
+				~COMPUTE_RESOURCE_LIMITS__FORCE_SIMD_DIST_MASK;
+	}
+
 	q->is_active = QUEUE_IS_ACTIVE(*q);
 }
 

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

@@ -532,6 +532,7 @@ struct queue_properties {
 enum mqd_update_flag {
 	UPDATE_FLAG_DBG_WA_ENABLE = 1,
 	UPDATE_FLAG_DBG_WA_DISABLE = 2,
+	UPDATE_FLAG_IS_GWS = 4, /* quirk for gfx9 IP */
 };
 
 struct mqd_update_info {

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

@@ -95,6 +95,7 @@ void kfd_process_dequeue_from_device(struct kfd_process_device *pdd)
 int pqm_set_gws(struct process_queue_manager *pqm, unsigned int qid,
 			void *gws)
 {
+	struct mqd_update_info minfo = {0};
 	struct kfd_node *dev = NULL;
 	struct process_queue_node *pqn;
 	struct kfd_process_device *pdd;
@@ -146,9 +147,10 @@ int pqm_set_gws(struct process_queue_manager *pqm, unsigned int qid,
 	}
 
 	pdd->qpd.num_gws = gws ? dev->adev->gds.gws_size : 0;
+	minfo.update_flag = gws ? UPDATE_FLAG_IS_GWS : 0;
 
 	return pqn->q->device->dqm->ops.update_queue(pqn->q->device->dqm,
-							pqn->q, NULL);
+							pqn->q, &minfo);
 }
 
 void kfd_process_dequeue_from_all_devices(struct kfd_process *p)