Xref: news2.mv.net comp.os.msdos.djgpp:1072 From: "William D. Kirby" Newsgroups: comp.os.msdos.djgpp Subject: Re: v2: optimization & floating point Date: 13 Feb 1996 10:55:10 GMT Organization: Netcom Lines: 36 Message-ID: <4fpqme$mse@ixnews6.ix.netcom.com> References: <4fo7dm$1fa AT news1 DOT usa DOT pipeline DOT com> NNTP-Posting-Host: ix-dc6-24.ix.netcom.com To: djgpp AT delorie DOT com DJ-Gateway: from newsgroup comp.os.msdos.djgpp knightofni AT usa DOT pipeline DOT com (The Knight Who Says Ni!) wrote: >If I compile with -O, the program will crash and say: >Exiting due to signal SIGFPE >Floating Point exception at eip=00001729 > >Is this a problem with v2.0, or do I need to use another command-line >option to use floating point with -O? > > > _______________________________ >/ \ The Knight Who Says "Ni!" \ >\_| | > | I'm not nearly as think | > | as you confused I am. | > | ___________________________|_ > \_/____________________________/ > I have been experiencing the same problems with DJGPP V2.0, and have had to always use no optimization. I have been trying to isolate the optimization option that causes the problem, but no luck so far. I have even run thru the DJGPP test programs for libc with optomization, and didn't get any errors. The problems of floating point exceptions and slower running code occured with the change from GCC v2.63 to v2.7.1 & 2.7.2. I think the problem is the original GCC source. -- William D. Kirby Voice: (703) 273-0005 Consultant Email: wdkirby AT ix DOT netcom DOT com (primary) 3527 Cornell Road or 70711 DOT 2407 AT compuserve DOT com (secondary) Fairfax, VA 22030-1813