Find cached display items directly during painting

third_party/WebKit/Source/core/paint/README.md

 # `Source/core/paint`
 
 This directory contains implementation of painters of layout objects. It covers
 the following document lifecycle states:
 
 *   PaintInvalidation (`InPaintInvalidation` and `PaintInvalidationClean`)
 *   PrePaint (`InPrePaint` and `PrePaintClean`)
 *   Paint (`InPaint` and `PaintClean`)
 
 ## Glossaries
@@ skipping 197 matching lines @@
     it's implemented.
 
 ## Paint result caching
 
 `PaintController` holds the previous painting result as a cache of display items.
 If some painter would generate results same as those of the previous painting,
 we'll skip the painting and reuse the display items from cache.
 
 ### Display item caching
 
-We'll create a `CachedDisplayItem` when a painter would create a `DrawingDisplayItem` exactly
-the same as the display item created in the previous painting. After painting, `PaintController`
-will replace `CachedDisplayItem` with the corresponding display item from the cache.
+When a painter would create a `DrawingDisplayItem` exactly the same as the display item
+created in the previous painting, we'll reuse the previous one instead of repainting it.
 
 ### Subsequence caching
 
 When possible, we enclose the display items that `PaintLayerPainter::paintContents()` generates
 (including display items generated by sublayers) in a pair of `BeginSubsequence/EndSubsequence`
 display items.
 
-In a subsequence paint, if the layer would generate exactly the same display items, we'll simply
-output a `CachedSubsequence` display item in place of the display items, and skip all paintings
-of the layer and its descendants in painting order. After painting, `PaintController` will
-replace `CacheSubsequence` with cached display items created in previous paintings.
+In a subsequence paint, if the layer would generate exactly the same display items, we'll get
+the whole subsequence from the cache instead of repainting them.
 
 There are many conditions affecting
 *   whether we need to generate subsequence for a PaintLayer;
 *   whether we can use cached subsequence for a PaintLayer.
 See `shouldCreateSubsequence()` and `shouldRepaintSubsequence()` in `PaintLayerPainter.cpp` for
 the conditions.
 
 ## Empty paint phase optimization
 
 During painting, we walk the layout tree multiple times for multiple paint phases. Sometimes
@@ skipping 12 matching lines @@
 
 When layer structure changes, and we are not invalidate paint of the changed subtree,
 we need to manually update the `needsPaintPhaseXXX` flags. For example, if an object changes
 style and creates a self-painting-layer, we copy the flags from its containing self-painting
 layer to this layer, assuming that this layer needs all paint phases that its container
 self-painting layer needs.
 
 We could update the `needsPaintPhaseXXX` flags in a separate tree walk, but that would regress
 performance of the first paint. For slimming paint v2, we can update the flags during the
 pre-painting tree walk to simplify the logics.