| Index: cc/trees/draw_property_utils.cc
|
| diff --git a/cc/trees/draw_property_utils.cc b/cc/trees/draw_property_utils.cc
|
| index 39456bc83bad09fa3d57a8c91b645147a4f29e5d..876645e1eed0e19221621dc02c8ef96682953546 100644
|
| --- a/cc/trees/draw_property_utils.cc
|
| +++ b/cc/trees/draw_property_utils.cc
|
| @@ -57,7 +57,6 @@ void CalculateVisibleRects(const std::vector<LayerType*>& visible_layer_list,
|
| success = transform_tree.ComputeTransformWithDestinationSublayerScale(
|
| clip_transform_node->id, target_node->id, &clip_to_target);
|
| }
|
| -
|
| if (target_node->id > clip_node->data.transform_id) {
|
| if (!success) {
|
| DCHECK(target_node->data.to_screen_is_animated);
|
| @@ -390,6 +389,7 @@ void ComputeClips(ClipTree* clip_tree, const TransformTree& transform_tree) {
|
| // clip adjusted due to intersecting with an ancestor clip.
|
| const bool is_clipped = clip_node->parent_id > 0;
|
| if (!is_clipped) {
|
| + DCHECK(!clip_node->data.inherit_parent_target_space_clip);
|
| clip_node->data.combined_clip = clip_node->data.clip;
|
| continue;
|
| }
|
| @@ -411,6 +411,8 @@ void ComputeClips(ClipTree* clip_tree, const TransformTree& transform_tree) {
|
| gfx::Transform parent_to_target;
|
| gfx::Transform clip_to_target;
|
| gfx::Transform target_to_clip;
|
| + gfx::Transform parent_to_transform_target;
|
| + gfx::Transform transform_target_to_target;
|
|
|
| const bool target_is_root_surface = clip_node->data.target_id == 1;
|
| // When the target is the root surface, we need to include the root
|
| @@ -419,8 +421,22 @@ void ComputeClips(ClipTree* clip_tree, const TransformTree& transform_tree) {
|
| target_is_root_surface ? 0 : clip_node->data.target_id;
|
|
|
| bool success = true;
|
| - if (parent_transform_node->data.content_target_id ==
|
| - clip_node->data.target_id) {
|
| + // When render surface applies clip, we need the clip from the target's
|
| + // target space. But, as the combined clip is in parent clip's target
|
| + // space, we need to first transform it from parent's target space to
|
| + // target's target space.
|
| + if (clip_node->data.inherit_parent_target_space_clip) {
|
| + success &= transform_tree.ComputeTransformWithDestinationSublayerScale(
|
| + parent_transform_node->id, transform_node->data.target_id,
|
| + &parent_to_transform_target);
|
| + success &= transform_tree.ComputeTransformWithSourceSublayerScale(
|
| + transform_node->data.target_id, target_id,
|
| + &transform_target_to_target);
|
| + transform_target_to_target.matrix().postScale(
|
| + transform_node->data.sublayer_scale.x(),
|
| + transform_node->data.sublayer_scale.y(), 1.0);
|
| + } else if (parent_transform_node->data.content_target_id ==
|
| + clip_node->data.target_id) {
|
| parent_to_target = parent_transform_node->data.to_target;
|
| } else {
|
| success &= transform_tree.ComputeTransformWithDestinationSublayerScale(
|
| @@ -444,25 +460,51 @@ void ComputeClips(ClipTree* clip_tree, const TransformTree& transform_tree) {
|
| // If we can't compute a transform, it's because we had to use the inverse
|
| // of a singular transform. We won't draw in this case, so there's no need
|
| // to compute clips.
|
| - if (!success)
|
| + if (!success) {
|
| continue;
|
| + }
|
|
|
| // In order to intersect with as small a rect as possible, we do a
|
| // preliminary clip in target space so that when we project back, there's
|
| // less likelihood of intersecting the view plane.
|
| - gfx::RectF inherited_clip_in_target_space = MathUtil::MapClippedRect(
|
| - parent_to_target, parent_clip_node->data.combined_clip);
|
| + gfx::RectF inherited_clip_in_target_space;
|
| + if (clip_node->data.inherit_parent_target_space_clip) {
|
| + gfx::RectF combined_clip_in_transform_target_space;
|
| + if (parent_transform_node->id > transform_node->data.target_id)
|
| + combined_clip_in_transform_target_space = MathUtil::MapClippedRect(
|
| + parent_to_transform_target, parent_clip_node->data.combined_clip);
|
| + else
|
| + combined_clip_in_transform_target_space = MathUtil::ProjectClippedRect(
|
| + parent_to_transform_target, parent_clip_node->data.combined_clip);
|
| + inherited_clip_in_target_space = MathUtil::ProjectClippedRect(
|
| + transform_target_to_target, combined_clip_in_transform_target_space);
|
| + } else if (parent_transform_node->id > target_id) {
|
| + inherited_clip_in_target_space = MathUtil::MapClippedRect(
|
| + parent_to_target, parent_clip_node->data.combined_clip);
|
| + } else {
|
| + inherited_clip_in_target_space = MathUtil::ProjectClippedRect(
|
| + parent_to_target, parent_clip_node->data.combined_clip);
|
| + }
|
|
|
| - gfx::RectF clip_in_target_space =
|
| - MathUtil::MapClippedRect(clip_to_target, clip_node->data.clip);
|
| + // When render surface inherits its parent target space clip, the layer
|
| + // that created the clip node doesn't apply any clip. So, we shouldn't clip
|
| + // using the clip value stored in the clip node.
|
| + gfx::RectF intersected_in_target_space;
|
| + if (!clip_node->data.inherit_parent_target_space_clip) {
|
| + gfx::RectF clip_in_target_space =
|
| + MathUtil::MapClippedRect(clip_to_target, clip_node->data.clip);
|
|
|
| - gfx::RectF intersected_in_target_space = gfx::IntersectRects(
|
| - inherited_clip_in_target_space, clip_in_target_space);
|
| + intersected_in_target_space = gfx::IntersectRects(
|
| + inherited_clip_in_target_space, clip_in_target_space);
|
| + } else {
|
| + intersected_in_target_space = inherited_clip_in_target_space;
|
| + }
|
|
|
| clip_node->data.combined_clip = MathUtil::ProjectClippedRect(
|
| target_to_clip, intersected_in_target_space);
|
|
|
| - clip_node->data.combined_clip.Intersect(clip_node->data.clip);
|
| + if (!clip_node->data.inherit_parent_target_space_clip)
|
| + clip_node->data.combined_clip.Intersect(clip_node->data.clip);
|
| }
|
| clip_tree->set_needs_update(false);
|
| }
|
|
|