https://gcc.gnu.org/bugzilla/show_bug.cgi?id=113576
--- Comment #39 from Hongtao Liu <liuhongt at gcc dot gnu.org> --- > > the question is whether that matches the semantics of GIMPLE (the padding > > is inverted, too), whether it invokes undefined behavior (don't do it - it > > seems for people using intrinsics that's what it is?) For the intrinisc, the instructions only care about lower bits, so it's not big issue? And it sounds like similar issue as _BitInt(4)/_BitInt(2), I assume there're garbage in the upper bits.
