[Debian-ppc64-devel] Call For Help - Please support the ppc64 architecture

Sven Luther sven.luther@wanadoo.fr
Thu, 17 Mar 2005 13:41:00 +0100


On Thu, Mar 17, 2005 at 10:46:31AM +0100, Cajus Pollmeier wrote:
> Am Donnerstag, 17. M=E4rz 2005 09:44 schrieb Sven Luther:
> > On Thu, Mar 17, 2005 at 09:46:36AM +1100, Benjamin Herrenschmidt wrot=
e:
> > > On Wed, 2005-03-16 at 20:27 +0100, Andreas Jochens wrote:
> > > > Hello,
> > > >
> > > > This is a call for help from the 'ppc64' porters.
> > > >
> > > > On 05-Mar-14 16:14, Martin Michlmayr wrote:
> > > > > Also, as with the amd64 port, there is disagreement about the n=
ame.
> > > > > While ppc64 would be nicer and in line with the LSB, our curren=
t
> > > > > PowerPC port is called powerpc and therefore it would make more=
 sense
> > > > > to call the 64 bit port powerpc64.
> > > >
> > > > There has been a decision of the Debian Technical Committee conce=
rning
> > > > the name of the amd64 port which basically says that the porting =
team
> > > > should decide on the architecture name generally (see [1]).
> > > >
> > > > The ppc64 porters decided to use the name 'ppc64' as the package
> > > > name a few month ago.
> > > >
> > > > .../...
> > >
> > > It's a fully 64 bits setup as it seems ? That is rather inefficient=
.
> > >
> > > Have we any proper way of doing multiarch setups ? The "proper" way=
 to
> > > do ppc64 is to have both archs libs and 32 bits userland for most
> > > things, as ppc64 native code is slightly slower.
> > >
> > > I have repeated that over and over again but it seems I have been
> > > ignored so far...
> >
> > Not ignored, there is an effort, fully orthogonal to this pure-64 one=
, to
> > get ppc64 biarch going. We are somewhat stopped by the work needed on=
 the
> > sarge release, but it will happen in the close next time.
> >
> > Now, there is an interest on IBM's and IBM's customer part for gettin=
g
> > ppc64 support, and altough we have access to the augsbourg power5 box=
 (but
> > without virtual machine, so we can't really do kernel or installer te=
sts),
> > we don't have those ppc64 machine IBM mentioned could be made availab=
le,
> > which makes work on the kernel and installer part at least less possi=
ble.
>=20
> Not happening to match 100% here, but:

We are speaking on real machines with full hand-on access.=20

> As said. I have such a machine and I can port the (currently working wi=
th a=20
> ppc64 kernel only) installer for these machines to be pure ppc64. You c=
an=20
> even do kernel tests if you want to, but I've to restart the correspond=
ing=20
> LPAR for you, then.

Yep, which can become bothersome once we really get this work going.

> If you test the kernel: I've done patches to enable vscsis in Debian an=
d added=20
> the emulex driver for vanilla 2.6.11.x. Who may take care of this?

Please fill a bugreport against the kernel, or send them to me directly. =
I am
building ppc64 kernels right now, but i will tell you once i have somethi=
ng
you can test.

Friendly,

Sven Luther