fast/repaint/fixed-tranformed.html started failing: http://test-results.appspot.com/dashboards/flakiness_dashboard.html#showExpectations=true&tests=fast%2Frepaint%2Ffixed-tranformed.html Blame list: http://trac.webkit.org/log/?verbose=on&rev=119620&stop_rev=119617 It's unconceivable that any changes in this range could have caused this test to regress.
On Linux, it looks like flakiness. On Windows, it looks like a 1px change in the scrollbars. Is that not expected from http://trac.webkit.org/changeset/119617/ ? Here's the diff: http://build.chromium.org/f/chromium/layout_test_results/Webkit_Win7/results/layout-test-results/fast/repaint/fixed-tranformed-diff.png
(In reply to comment #1) > On Linux, it looks like flakiness. On Windows, it looks like a 1px change in the scrollbars. Is that not expected from http://trac.webkit.org/changeset/119617/ ? > > Here's the diff: http://build.chromium.org/f/chromium/layout_test_results/Webkit_Win7/results/layout-test-results/fast/repaint/fixed-tranformed-diff.png Oh yeah, so can probably rebaseline Windows.
Rebaselined in http://trac.webkit.org/changeset/119863 . According to the flakiness dashboard, this test was passing on Mac and Linux, so I removed the line from TestExpectations.
Actually, it's flaky on Linux, but there is a separate line for that in TestExpectations.