--zhXaljGHf11kAtnf
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

* Denis Lagno (dlagno@smtp.ru) wrote:
> > > The most likely answer is that the NetBSD kernel source code accesses=
 memory=20
> > > not entirely in a type-safe fashion, and higher optimization levels w=
ill=20
> > > expose these problems because the compiler's assumptions about which
> > > variables can be changed via pointer, structure, or union member
> > > dereferencing are not correct.
> >=20
> > I don't think so. Kernels compile with "-march=3Dp4" and "-O2" work fine
> > for me on two differnt systems. My guess is a code generation bug in GC=
C,
> > IIRC at least "-march=3Dpentium" is broken.
>=20
> Seconded.  I have booting problems on Transmeta TM-5800 based laptop.
> -fno-strict-aliasing or -fno-rename-registers do not help.
> Working kernel results only when -march=3Dpentium removed.

I own a desknote Laptop with a GigaPro CPU (TH5600) running a GENERIC
Kernel and Userland with 2.0G

> Offtopic question: is there any way to run TM-5800 in native mode?

No, I did some research in that CPU 1,5a ago and found no way.=20
You have to bypass that i386 morphcode, and that would be to hard.
But you could inject Alpha or sun64 morphcode to let your transmeta act
like a Alpha or Sparc ;-)
I Don't know if an actual Crusoe is able to run w/o morphcode, because
it is a very RISCy CPU.

--=20
Und die Liebe
Eine Illusion
Du tanzt im Licht der Zeit
Du tanzt in Eitelkeit                        - LACRIMOSA

--zhXaljGHf11kAtnf
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (NetBSD)

iD8DBQFBKfkGEfTEHrP7rjMRArmJAJ0ZVs8+PV/5Cfzu8Cddr+3bKrIFvACdGSVC
nau66YWmcLPZb7ZkNVBh5yc=
=L7qr
-----END PGP SIGNATURE-----

--zhXaljGHf11kAtnf--