egcs 2.91.66 ix86 problems...

[Geoff Keating]

> Mailing-List: contact libc-hacker-help@sourceware.cygnus.com; run by ezmlm
> Date: Tue, 15 Jun 1999 19:09:26 -0400 (EDT)
> From: Cristian Gafton <gafton@redhat.com>
> 
> 
> I am going nuts over this issue, maybe someone can help me see the light
> here:
...
> So after reading this thing one clearly gets the impression the saying
> -mcpu=i486 will affect the scheduler only, while egcs will still generate
> code that will run on i386, and -march=i486 will actually enable
> i486-specific instructions.
> 
> Well, the cpp does not concur to this idea and because of this the glibc
> headers that are testing for __i486__ (like the ones that define bswap32
> and bswap16) are getting confused.

It would be easy for it to be confused; on powerpc, `-mcpu=601' means
produce code that runs only on 601, and `-mtune=601' means write code
that runs on any architecture but is scheduled (and so on) for the 601.

This is in fact how it works for cpp, too; _ARCH_PWR is defined with
-mcpu=601, but not defined with -mtune=601.

-- 
Geoffrey Keating <geoffk@ozemail.com.au>