https://gcc.gnu.org/bugzilla/show_bug.cgi?id=122956
--- Comment #6 from Andrew Pinski <pinskia at gcc dot gnu.org> --- (In reply to Andrew Pinski from comment #5) > (In reply to Hongtao Liu from comment #4) > > (In reply to Andrew Pinski from comment #3) > > > (In reply to Andrew Pinski from comment #2) > > > > On a related subject did spec next fix llvm code up for its undefined > > > > code? > > > > > > So at least upstream didn't. This would mean spec (again) still depends on > > > undefined code. > > > > Which part is UD? > > See https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113341 and the linked bugs > there. Actually there is another few about it. Basically llvm depends on values stored in a pointer which then calls operator new on it.
