"William M. Quarles" <quarlewm@xxxxxxx> writes:
> Hi,
>
> I heard this gnasty rumor attached below that says that GCC
> can't optimize any differently between any of the P6 or later
> processors, so that -march=pentiumpro and -march=pentium2 for
-march=pentium2 allows generation of mmx instrctions,
-march=pentiumpro does not. (I don't know if the optimizer makes
much use of mmx however.) I believe -march=pentium2 also implies
different parameters to the scheduler.
See
http://gcc.gnu.org/onlinedocs/gcc-3.3/gcc/i386-and-x86-64-Options.html#i386%20and%20x86-64%20Options
for more info on these options.
> example will always compile the same. I have not been able to
> find anything in the GCC documentation, buglist, or
> limitations list that suggests that this is the case. Is what
> this developer telling me true?
[snip]
They will often result in the same code, but not always. They do
affect use of the x86 builtins
see:
http://gcc.gnu.org/onlinedocs/gcc-3.3/gcc/X86-Built-in-Functions.html#X86%20Built-in%20Functions
