80331 2012-03-05 14:52:40 -0800 [chromium] Refactor CCLayerTreeHostCommon: untangle drawableContentRect, part 1 2012-07-16 09:09:13 -0700 1 1 1 Unclassified WebKit Layout and Rendering 528+ (Nightly build) Unspecified Unspecified RESOLVED DUPLICATE 80622 P2 Normal --- 1 shawnsingh shawnsingh danakj enne oldest_to_newest 571068 0 shawnsingh 2012-03-05 14:52:40 -0800 The computation and usage of drawableContentRect is one of the most confusing aspects of calculateDrawTransformsAndVisibility: (1) before recursion, drawableContentRect is initalized to the clipRect (or, if no masksToBounds, the layer's boundaries) (2) after recursion, drawableContentRect is expanded to include layers, and is then re-clipped if masksToBounds is true. (3) after all that, the clipRect is re-initialized to the drawableContentRect ... ?! (4) drawableContentRect should not be a layer property; it is not actually used outside calculateDrawTransformsAndVisibility; it is only used to initialize the target renderSurface's contentRect. Cleaning this up safely will take multiple patches, this first patch will try to remove some redundancy in the appropriate if statements, before trying to clean up the clipRect/drawableContentRect knot described above. 587909 1 danakj 2012-03-26 11:00:11 -0700 This may or may not be the place to fix the clipRect + replica situation. But the clipRect for both the surface is wrong when the surface has a replica (it includes the replica and possibly more), and we don't have a clipRect for just the replica. 670384 2 shawnsingh 2012-07-16 09:09:13 -0700 *** This bug has been marked as a duplicate of bug 80622 ***