To avoid sizing and repaint issues we should layout replaced elements on pixel bounds. We already ensure that replaced elements are sized in full pixels and that they are painted on pixel bounds. By also ensuring that they are placed on pixel bounds we avoid pixel having the size be expanded by pixel snapping and repainting/invalidation rect issues when scrolling. Downstream chromium bug: http://code.google.com/p/chromium/issues/detail?id=133792
Created attachment 150343 [details] Patch
Comment on attachment 150343 [details] Patch Attachment 150343 [details] did not pass chromium-ews (chromium-xvfb): Output: http://queues.webkit.org/results/13131010 New failing tests: platform/chromium/compositing/layout-width-change.html platform/chromium/compositing/render-surface-alpha-blending.html platform/chromium/compositing/tiny-layer-rotated.html platform/chromium/compositing/huge-layer-rotated.html fast/loader/loadInProgress.html platform/chromium/compositing/3d-corners.html platform/chromium/compositing/video-frame-size-change.html platform/chromium/compositing/perpendicular-layer-sorting.html fast/loader/unload-form-post-about-blank.html platform/chromium/compositing/child-layer-3d-sorting.html platform/chromium/compositing/lost-compositor-context-permanently.html platform/chromium/compositing/lost-compositor-context-with-video.html http/tests/xmlhttprequest/zero-length-response.html http/tests/security/script-crossorigin-loads-correctly.html fast/repaint/block-selection-gap-in-composited-layer.html media/media-document-audio-repaint.html platform/chromium/compositing/backface-visibility-transformed.html platform/chromium/compositing/lost-compositor-context-with-rendersurface.html platform/chromium/compositing/lost-compositor-context.html inspector/timeline/timeline-receive-response-event.html platform/chromium/compositing/img-layer-grow.html platform/chromium/compositing/lost-compositor-context-twice.html
Created attachment 150348 [details] Archive of layout-test-results from ec2-cr-linux-03 The attached test failures were seen while running run-webkit-tests on the chromium-ews. Bot: ec2-cr-linux-03 Port: <class 'webkitpy.common.config.ports.ChromiumXVFBPort'> Platform: Linux-2.6.35-28-virtual-x86_64-with-Ubuntu-10.10-maverick
Comment on attachment 150343 [details] Patch Looks reasonable, but clearly you have some expecatations to update yet. :)
The chromium ews bot test results don't match what I'm seeing locally. Looking into it now.
> The chromium ews bot test results don't match what I'm seeing locally. Looking into it now. We just switched hardware for the bots, so I wouldn't be surprised if it was having some trouble.
Created attachment 150495 [details] Patch for landing
Comment on attachment 150495 [details] Patch for landing Clearing flags on attachment: 150495 Committed r121722: <http://trac.webkit.org/changeset/121722>
All reviewed patches have been landed. Closing bug.