[chromium] Don't set visible layer rect in CCLayerTreeHost paint iteration
Created attachment 147120 [details] Patch
Comment on attachment 147120 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=147120&action=review > Source/WebCore/platform/graphics/chromium/cc/CCLayerTreeHostCommon.cpp:840 > } else if (it.representsContributingRenderSurface()) { Can the root layer have a mask? (it won't be contributing)
Created attachment 147126 [details] Handle root mask
Comment on attachment 147126 [details] Handle root mask View in context: https://bugs.webkit.org/attachment.cgi?id=147126&action=review > Source/WebCore/platform/graphics/chromium/cc/CCLayerTreeHostCommon.cpp:839 > + if (it.representsTargetRenderSurface()) { > + LayerType* maskLayer = it->maskLayer(); > + if (maskLayer) > + maskLayer->setVisibleLayerRect(IntRect(IntPoint(), it->contentBounds())); > + LayerType* replicaMaskLayer = it->replicaLayer() ? it->replicaLayer()->maskLayer() : 0; > + if (replicaMaskLayer) > + replicaMaskLayer->setVisibleLayerRect(IntRect(IntPoint(), it->contentBounds())); Cool, that's better. What do you think of a continue at the end of this block? The If/else situation is a bit confusing, and that might clear it up, letting us drop the else and un-nest that block below. Either way, the change looks correct to me.
Created attachment 147132 [details] Conditional cleanup
Comment on attachment 147132 [details] Conditional cleanup This LGTM, thanks for the changes. The logic is changed a bit, and we won't have a visibleLayerRect set on layers that own a surface but don't drawContents themselves, but that is probably just fine.
Comment on attachment 147132 [details] Conditional cleanup Nice!
Comment on attachment 147132 [details] Conditional cleanup Clearing flags on attachment: 147132 Committed r120129: <http://trac.webkit.org/changeset/120129>
All reviewed patches have been landed. Closing bug.