--gj572EiMnwbLXET9
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

Hi,

On Thu, Sep 26, 2002 at 12:22:28AM +0200, Klaus Heinz wrote:

>  [ need of busspace.*_4 ]
>=20
> > Up to know, we never needed this, as we never had anything with 32bit
> > registers in a busified driver.
>=20
> Hmm, maybe those 32 bits made me think about cardbus...PCMCIA can only
> handle 16 bits.
>=20
> > e.g., PCMCIA can't do this trick, and the ZII devices we're handling=20
> > through busXXX() are 8 or 16bit.
>=20
> Obviously it was added in v1.41 of dev/ic/smc91cxx.c. Every occurrence
> of the bus_space.*_4 functions is in the 'else' branch of a test for
> the MD flag SMC_FLAGS_32BIT_READ which will hopefully never be set on
> an Amiga.
>=20
> Should we add more 'fake 32-bit macros' in amiga/include/bus.h ?

Please not. They should probably be split up into two 16bit accesses in a
meaningful way - for LE busses low order first.

If you add a fake access, it should panic() the machine, at least for=20
a kernel with DEBUG set.

Regards,
	-is

--gj572EiMnwbLXET9
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (NetBSD)
Comment: For info see http://www.gnupg.org

iD8DBQE9k2icPCRcZ/VMtk4RAqN/AJ9yuy5MWgGaLOBHi4P/9h+lxHObFQCfRz0N
3dl+WvZLamIBtD+zuA2ugjM=
=F9zf
-----END PGP SIGNATURE-----

--gj572EiMnwbLXET9--