Re: Why does config.guess come up with such wacky stuff?

[Robert Lipe <robertlipe at usa dot net>]

This is bordering between an autoconf and a gcc discussion.  Let's
decide if it needs fixed on this list and then move the "how" to
autoconf.

Alexandre Oliva wrote:
> On Dec  1, 1999, "John Hughes" <john@Calva.COM> wrote:
> 
> > It's a pity that the autoconf version looses the CPU type information for
> > UnixWare 7, just forcing everything to i586:
> 
> What's the difference?  Gcc won't change default optimizations for x86
> depending on the build machine, AFAIK.

I believe it does.  The code around line 3560 of configure.in handles
this.  So now if you're on a 386 and type 'configure && make', GCC
will presumably emit opcodes that your computer won't understand and
generated binaries will crash.  If you're on a P-II or P-III and do
the same, you're leaving clock cycles on the table becuase GCC is now
"dumbed down" for a P5.  Neither of these are good.

I've added the guy that I believe to be the author of this specific
change to this discussion.  Hops, I feel pretty strongly that
config.guess should cater to the '/path/to/configure && make && make
install' user and that creators of binary kits should be responsible
for overriding the triple if they so desire.  Without fretting over the
details right now, would you support something like this in the master
copy of config.guess so that GCC can use the master copy?

case `uname -X | egrep '^Machine` in
        *i80486)   UNAME_MACHINE=i486;;
        *Pentium$) UNAME_MACHINE=i586;;
        *PentII*)  UNAME_MACHINE=i686;;
esac