Jonathan Watt and Michal Zalewski independently reported timing dependent testcases that trigger crashes at the same place during text display. We have seen no demonstration that these crashes could be reliably exploited, but they do show evidence of memory corruption so we presume they could be.