http://gcc.gnu.org/bugzilla/show_bug.cgi?id=59265
--- Comment #19 from Markus Trippelsdorf <trippels at gcc dot gnu.org> --- (In reply to Jan Hubicka from comment #18) > > Just double-checked. And yes, unfortunately it's 100% reproducible. > > Can not think of much of reason for this. How much off the index is? Just > slightly or is it completely random number? Off by one: imple_stmt_max_uid (fn) < cedge->lto_stmt_uid) 9 < 10
