fast/viewport/ios/width-is-device-width-overflowing.html is now a flaky failure, causing huge EWS slowdown and false positives. Depending on the bot, it either fails most of the time, or passes most of the time.
Not sure if other related tests are affected.
I removed this from TE yesterday hoping that it was better, but I guess not. Will look into it.
Looks like nothing happened yet.
I've put back expectations for these three related tests in http://trac.webkit.org/r205013, let's keep tracking the problem in bug 153110.
*** This bug has been marked as a duplicate of bug 153110 ***
I could not reproduce the flakiness locally.