------- Comment #5 from rguenther at suse dot de 2008-05-29 10:59 ------- Subject: Re: [4.1/4.2/4.3/4.4 Regression] ICE in simplify_subreg
On Thu, 29 May 2008, ubizjak at gmail dot com wrote: > ------- Comment #4 from ubizjak at gmail dot com 2008-05-29 10:23 ------- > The problem is, that we enter simplify_gen_subreg from operand_subword with: > > op: > (reg:QI 61) > > outermode: word_mode (SImode) > innermode: DImode > > This triggers assert in simplify_subreg: > > gcc_assert (GET_MODE (op) == innermode > || GET_MODE (op) == VOIDmode); I believe this goes wrong on the tree level somewhere, as we have a TRUTH_NOT_EXPR of long long type with a boolean type argument (NE_EXPR). But I cannot see where this broken typed trees are built. Richard. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=36362