Bug 1441308 - Fix picture-caching interaction with doc spitting r=gw

This is a stab at what the correct approach to this should be. It
seems that we should be using the window size here and not the
screen_rect, as the screen_rect is not used to offset what we
normally draw, but instead generally for scissoring(?). The end
result is if we use an offset screen_rect, we end up applying
the offset of the chrome area twice, once because the document's
screen rect is offset, and once because of the tile.world_rect
offset.

Depends on D20696

Differential Revision: https://phabricator.services.mozilla.com/D20698

[wrupdater] From https://hg.mozilla.org/mozilla-central/rev/48bf7d6e2ea7be2eff20cafdbb4c79dfd3706e8e

Author: squarewave

webrender/src/frame_builder.rs

@@ -475,8 +475,11 @@ impl FrameBuilder {
             .take_render_tasks();
 
         let root_render_task = RenderTask::new_picture(
-            RenderTaskLocation::Fixed(self.screen_rect.to_i32()),
-            self.screen_rect.size.to_f32(),
+            // The rect here is the whole window. If we were to use the screen_rect,
+            // any offset in that rect would doubly apply for cached pictures.
+            RenderTaskLocation::Fixed(DeviceIntRect::new(DeviceIntPoint::zero(),
+                                                         self.window_size).to_i32()),
+            self.window_size.to_f32(),
             self.root_pic_index,
             DeviceIntPoint::zero(),
             child_tasks,