https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66780
--- Comment #4 from Kazumoto Kojima <kkojima at gcc dot gnu.org> --- (In reply to John Paul Adrian Glaubitz from comment #2) > Could this also be the cause for PR66312 and PR66563 after all? My 2 cents. I've just revert the problematic part on trunk after usual test. I'll revert it on 4.9 when the usual test is done and to 5 when the branch reopens.
