http://gcc.gnu.org/bugzilla/show_bug.cgi?id=59700
--- Comment #16 from Dominique d'Humieres <dominiq at lps dot ens.fr> --- I agree with Jerry that the errors with logical can be fixed later (AFAICT they are not a regression up to 4.3.1). I also think the error read(fd, *, err=31, iomsg=msg) i1, x2, a, x1 31 if (msg /= 'Bad repeat count in item 3 of list input') call abort is bogus (IIRC repeat count needs a '*').
