[PATCH v1 10/20] drm/amdgpu: prepare amdgpu_fill_buffer to use N entities

[Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>]

We want to use multiple entities for clears to enable parallelism.
To achieve this, let amdgpu_fill_buffer select the entity to use
when it's possible by passing a NULL entity.

We can also simplify the signature and remove the resv param:
amdgpu_move_blit is the only caller that doesn't use it, and it's
also the only caller that needs to specify an entity.

Signed-off-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 11 +----
 drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c    | 54 ++++++++++++++++++----
 drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h    |  1 -
 3 files changed, 47 insertions(+), 19 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
index be3532134e46..4a69324bb730 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
@@ -1287,7 +1287,6 @@ void amdgpu_bo_move_notify(struct ttm_buffer_object *bo,
 void amdgpu_bo_release_notify(struct ttm_buffer_object *bo)
 {
 	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->bdev);
-	struct dma_fence *fence = NULL;
 	struct amdgpu_bo *abo;
 	int r;
 
@@ -1317,18 +1316,12 @@ void amdgpu_bo_release_notify(struct ttm_buffer_object *bo)
 	    adev->in_suspend || drm_dev_is_unplugged(adev_to_drm(adev)))
 		goto out;
 
-	r = dma_resv_reserve_fences(&bo->base._resv, 1);
-	if (r)
-		goto out;
-
-	r = amdgpu_fill_buffer(&adev->mman.clear_entities[0], abo, 0, &bo->base._resv,
-			       &fence, NULL, AMDGPU_KERNEL_JOB_ID_CLEAR_ON_RELEASE);
+	r = amdgpu_fill_buffer(NULL, abo, 0, NULL,
+			       NULL, AMDGPU_KERNEL_JOB_ID_CLEAR_ON_RELEASE);
 	if (WARN_ON(r))
 		goto out;
 
 	amdgpu_vram_mgr_set_cleared(bo->resource);
-	dma_resv_add_fence(&bo->base._resv, fence, DMA_RESV_USAGE_KERNEL);
-	dma_fence_put(fence);
 
 out:
 	dma_resv_unlock(&bo->base._resv);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
index b94ac16c785b..c357a6d9763a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
@@ -184,7 +184,8 @@ static int amdgpu_ttm_map_buffer(struct drm_sched_entity *entity,
 				 struct amdgpu_res_cursor *mm_cur,
 				 unsigned int window, struct amdgpu_ring *ring,
 				 bool tmz, uint64_t *size, uint64_t *addr,
-				 struct dma_fence *dep)
+				 struct dma_fence *dep,
+				 struct dma_resv *resv)
 {
 	struct amdgpu_device *adev = ring->adev;
 	unsigned int offset, num_pages, num_dw, num_bytes;
@@ -239,6 +240,10 @@ static int amdgpu_ttm_map_buffer(struct drm_sched_entity *entity,
 	if (dep)
 		drm_sched_job_add_dependency(&job->base, dma_fence_get(dep));
 
+	if (resv)
+		drm_sched_job_add_resv_dependencies(&job->base, resv,
+						    DMA_RESV_USAGE_BOOKKEEP);
+
 	src_addr = num_dw * 4;
 	src_addr += job->ibs[0].gpu_addr;
 
@@ -332,14 +337,14 @@ static int amdgpu_ttm_copy_mem_to_mem(struct amdgpu_device *adev,
 		r = amdgpu_ttm_map_buffer(&entity->base,
 					  src->bo, src->mem, &src_mm,
 					  entity->gart_window_id0, ring, tmz, &cur_size, &from,
-					  NULL);
+					  NULL, NULL);
 		if (r)
 			goto error;
 
 		r = amdgpu_ttm_map_buffer(&entity->base,
 					  dst->bo, dst->mem, &dst_mm,
 					  entity->gart_window_id1, ring, tmz, &cur_size, &to,
-					  NULL);
+					  NULL, NULL);
 		if (r)
 			goto error;
 
@@ -422,7 +427,7 @@ static int amdgpu_move_blit(struct ttm_buffer_object *bo,
 		struct dma_fence *wipe_fence = NULL;
 
 		r = amdgpu_fill_buffer(&adev->mman.move_entities[0],
-				       abo, 0, NULL, &wipe_fence, fence,
+				       abo, 0, &wipe_fence, fence,
 				       AMDGPU_KERNEL_JOB_ID_MOVE_BLIT);
 		if (r) {
 			goto error;
@@ -2451,7 +2456,7 @@ int amdgpu_ttm_clear_buffer(struct amdgpu_bo *bo,
 		r = amdgpu_ttm_map_buffer(&entity->base,
 					  &bo->tbo, bo->tbo.resource, &cursor,
 					  entity->gart_window_id1, ring, false, &size, &addr,
-					  NULL);
+					  NULL, NULL);
 		if (r)
 			goto err;
 
@@ -2472,10 +2477,21 @@ int amdgpu_ttm_clear_buffer(struct amdgpu_bo *bo,
 	return r;
 }
 
+/**
+ * amdgpu_fill_buffer - fill a buffer with a given value
+ * @entity: optional entity to use. If NULL, the clearing entities will be
+ *          used to load-balance the partial clears
+ * @bo: the bo to fill
+ * @src_data: the value to set
+ * @f: optional out fence. If @entity is NULL, this must be NULL and the
+ *     fences from each partial clear will be added to the &dma_resv.
+ * @dependency: optional input dependency fence.
+ * @k_job_id: trace id
+ *
+ */
 int amdgpu_fill_buffer(struct amdgpu_ttm_entity *entity,
 		       struct amdgpu_bo *bo,
 		       uint32_t src_data,
-		       struct dma_resv *resv,
 		       struct dma_fence **f,
 		       struct dma_fence *dependency,
 		       u64 k_job_id)
@@ -2483,15 +2499,29 @@ int amdgpu_fill_buffer(struct amdgpu_ttm_entity *entity,
 	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
 	struct amdgpu_ring *ring = adev->mman.buffer_funcs_ring;
 	struct dma_fence *fence = NULL;
+	struct dma_resv *resv = NULL;
 	struct amdgpu_res_cursor dst;
 	int r;
 
+	/* The fences will be either added to the resv object or the last fence
+	 * will be returned to the caller. In the latter case, all fill jobs will
+	 * be executed on the same ring.
+	 */
+	WARN_ON_ONCE((entity && !f) || (!entity && f));
 	if (!adev->mman.buffer_funcs_enabled) {
 		dev_err(adev->dev,
 			"Trying to clear memory with ring turned off.\n");
 		return -EINVAL;
 	}
 
+	if (!entity) {
+		entity = &adev->mman.clear_entities[0];
+		resv = &bo->tbo.base._resv;
+		r = dma_resv_reserve_fences(resv, 1);
+		if (r)
+			return r;
+	}
+
 	amdgpu_res_first(bo->tbo.resource, 0, amdgpu_bo_size(bo), &dst);
 
 	mutex_lock(&entity->gart_window_lock);
@@ -2506,7 +2536,8 @@ int amdgpu_fill_buffer(struct amdgpu_ttm_entity *entity,
 					  &bo->tbo, bo->tbo.resource, &dst,
 					  entity->gart_window_id1, ring, false,
 					  &cur_size, &to,
-					  dependency);
+					  dependency,
+					  resv);
 		if (r)
 			goto error;
 
@@ -2516,8 +2547,13 @@ int amdgpu_fill_buffer(struct amdgpu_ttm_entity *entity,
 		if (r)
 			goto error;
 
-		dma_fence_put(fence);
-		fence = next;
+		if (resv) {
+			dma_resv_add_fence(resv, next, DMA_RESV_USAGE_KERNEL);
+			dma_fence_put(next);
+		} else {
+			dma_fence_put(fence);
+			fence = next;
+		}
 
 		amdgpu_res_next(&dst, cur_size);
 	}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
index 9dd2a76a5641..38df2b5b4bc7 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
@@ -184,7 +184,6 @@ int amdgpu_ttm_clear_buffer(struct amdgpu_bo *bo,
 int amdgpu_fill_buffer(struct amdgpu_ttm_entity *entity,
 		       struct amdgpu_bo *bo,
 		       uint32_t src_data,
-		       struct dma_resv *resv,
 		       struct dma_fence **f,
 		       struct dma_fence *dependency,
 		       u64 k_job_id);
-- 
2.43.0