--MGYHOYXEY6WxJCY8
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Mon, Feb 07, 2005 at 11:13:21AM +0100, Gunther Nikl wrote:
> On Sat, Feb 05, 2005 at 09:26:25PM +0100, Hubert Feyrer wrote:
> >=20
> > I stumbled across the following line in=20
> > src/sys/arch/amiga/conf/Makefile.amiga:
> >=20
> > 	.if empty(IDENT:M-DM68060)
> > 	CMACHFLAGS=3D     -m68020
> > 	.else
> > =3D=3D=3D>	CMACHFLAGS=3D     -m68060 -Wa,-m68030 -Wa,-m68851
> > 	.endif
> >=20
> > My question is, why bother with producing code for the 68030 when it's=
=20
> > known that the target is a 68060?
>=20
>   Assembler switches are different to compiler switches. With -Wa you
>   tell the assembler that it shall recognize certain instructions. Those
>   -Wa are nothing to worry about.

To be precise: as the assembler can't be told to recognize all CP/MMU
variants' MMU and cache syntaxes at once, and for some CPU/MMUs, the same
mnemonic translates to different instruction codes, we have to chose one
cpu/mmu with which we use symbolic instructions, and we're using .word
pseudo-instructions for all the others.

Regards,
	-is
--=20
seal your e-mail: http://www.gnupg.org/

--MGYHOYXEY6WxJCY8
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (NetBSD)

iD8DBQFCCP86N4tiz3B8hB0RAgXkAKDcVjD+eZbxLmVYka8ZTqb5MqbTYgCgtmxk
3RWyvlQtuUCbxjb/PTTJfuM=
=8DJc
-----END PGP SIGNATURE-----

--MGYHOYXEY6WxJCY8--