https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104901
--- Comment #9 from David Binderman <dcb314 at hotmail dot com> --- (In reply to Andreas Schwab from comment #7) > I think it would still be worthwhile to avoid the antipattern (v == c1 && v > == c2). +1 I haven't got a rs6000 box, but I suspect our old friend -Wlogical-op would find this problem also.