From: ao950 AT FreeNet DOT Carleton DOT CA (Paul Derbyshire) Newsgroups: comp.os.msdos.djgpp Subject: Re: pentium Date: 18 Feb 1997 06:02:16 GMT Organization: The National Capital FreeNet Lines: 32 Message-ID: <5ebgl8$lan@freenet-news.carleton.ca> References: <199702161216 DOT HAA10376 AT delorie DOT com> Reply-To: ao950 AT FreeNet DOT Carleton DOT CA (Paul Derbyshire) NNTP-Posting-Host: freenet2.carleton.ca To: djgpp AT delorie DOT com DJ-Gateway: from newsgroup comp.os.msdos.djgpp Dave Love (d DOT love AT dl DOT ac DOT uk) writes: >>>>>> "Martynas" == Martynas Kunigelis writes: > > Martynas> Do you mean that pentium optimized code does not run on > Martynas> 386/486? > > I don't know about that, but the `pgcc' hacks that I last saw actually > were even at least somewhat specific to the GCC front end language > (tough on Fortran, for instance). I've also been told that they are > 586-specific and not appropriate (as optimizations) for 686. 686??? > For the brave, a snapshot of the current gcc development version, > which has (different) [56]86 knowledge can be had from the GNU `win'32 > distribution at ftp.cygnus.com. [Other front ends than the C-based > ones may not build against it straight off; G77 won't, anyhow.] Why not include an option to build a program into a "super-executable" with three copies of the code in it, and at the start a routine to autodetect whether you have <586, 586, or 686 and branch to the right code, which is properly optimised for that chip? (Of course exe's would be up to three times as big, except for the 200k startup-code and stub.) -- .*. Where feelings are concerned, answers are rarely simple [GeneDeWeese] -() < When I go to the theater, I always go straight to the "bag and mix" `*' bulk candy section...because variety is the spice of life... [me] Paul Derbyshire ao950 AT freenet DOT carleton DOT ca, http://chat.carleton.ca/~pderbysh