Bug originally reported in http://code.google.com/p/chromium/issues/detail?id=136017 ... The assertion is being triggered due to anti-aliasing inflation that causes clipping. To our knowledge, this scenario of clipping is actually tolerable, and has not yet caused any known bugs. In addition to fixing the assert, I updated comments that needed to be worded better. Patch coming in a moment
Created attachment 152284 [details] Patch
Comment on attachment 152284 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=152284&action=review R=me, with comment nits. > Source/WebCore/platform/graphics/chromium/LayerRendererChromium.cpp:703 > + // (Note, however, that it may still have perspective components in the 2d parts of the matrix.) I don't know that it's important to add this superfluous comment about perspective here. Somebody modifying this code in the future doesn't need to be made aware of this extra detail just because this comment was wrong in the past. > Source/WebCore/platform/graphics/chromium/LayerRendererChromium.cpp:884 > + // (Note, however, that it may still have perspective components in the 2d parts of the matrix.) Same with this perspective comment. > Source/WebCore/platform/graphics/chromium/LayerRendererChromium.cpp:-885 > - ASSERT(!clipped); I do think you need a comment about why it's ok that this quad is clipped, since we assert on clipped pretty much everywhere else.
Created attachment 152288 [details] Patch for landing
Thanks for the review! I'm happy to tweak the comments more if they're not good enough yet, please feel free to cq- it if needed.
Comment on attachment 152288 [details] Patch for landing Clearing flags on attachment: 152288 Committed r122619: <http://trac.webkit.org/changeset/122619>
All reviewed patches have been landed. Closing bug.