Created attachment 170627 [details]
Repaint artifacts on Chromium.
Chromium bug: http://code.google.com/p/chromium/issues/detail?id=156597
The attached test exhibits repaint "ghosting" on Chromium.
Subpixel stroke widths trigger Skia's hairline optimization (treating the stroke width as == 1 and modulating the color components). Combined with subpixel positioning, this means that we may end up touching pixels outside the computed repaint rect.
Mike mentioned he's considering addressing this in Skia if he can find a way that doesn't impact the hairline optimization. In the meantime, we could inflate the repaint rect by .5 when USE(SKIA) to make sure we're covering the stroke.
Marked LayoutTest bugs, bugs with Chromium IDs, and some others as WontFix. Test failure bugs still are trackable via TestExpectations or disabled unit tests.