Path: utzoo!utgpu!water!watmath!clyde!bellcore!rutgers!caip.rutgers.edu!aramis.rutgers.edu!athos.rutgers.edu!hedrick From: hedrick@athos.rutgers.edu (Charles Hedrick) Newsgroups: comp.sys.pyramid Subject: Re: compiling GNU Emacs 18.51's data.c -OG -v on Pyramid OSx 4.0 Message-ID: Date: 23 Jul 88 22:57:10 GMT References: <18370@tut.cis.ohio-state.edu> Organization: Rutgers Univ., New Brunswick, N.J. Lines: 13 >Has anyone else observed this with any other optimizing compilers? Bugs in optimizing compilers are notoriously common. Technology must be improving, because Pyramid's and Sun's recent optimizers are fairly good. But the Sun 4.0 compiler on a Sun 2 loops when compiling one of the Gnu Emacs modules. (It would be interesting if it is the same one.) And of course C compilers for the Intel chips (both MS-DOS and ATT's compiler for System V/286) are notorious for causing trouble. I haven't seen a loop in the 286 compiler I use yet, but I have seen lots of cases where the cc top-level says /lib/optim blew up, so it is not using optimization, or where cc as a whole coredumps. I've even seen some cases where bad code is generated. That's quite rare on Sun and Pyramid compilers.