Comment 163 for bug 727620

I've tested the patched natty kernel and have over 30 boots without the evergreen_cp_resume oops, so the bisected patch indeed appears to be the right one. Does it explain the race?