lcs ethernet driver source

lcs ethernet driver source

[Arjan van de Ven]

Hi,

I tried to find the source for the lcs and qeth ethernet drivers for IBM's
S390 architecture, so I could recompile the kernel. However I seem to not be
able to find the source of these. Does anyone have a URL handy for this ?
(I wonder why these aren't in Linus or Alan's tree like most of the other
s390 patches)

Greetings,
   Arjan van de Ven

Re: lcs ethernet driver source

[Florian Weimer]

"Ulrich Weigand" <Ulrich.Weigand@de.ibm.com> writes:

> If you are running Linux native on the S/390 (not in a virtual machine)
> and have *only* physical LCS or QETH devices, you must use the binary-only
> modules; but this doesn't appear to be fundamentally different from other
> devices where only binary drivers exist ...

Well, it's unsual for such companies to be Patrons of the FSF.

SCNR.

-- 
Florian Weimer 	                  Florian.Weimer@RUS.Uni-Stuttgart.DE
University of Stuttgart           http://cert.uni-stuttgart.de/
RUS-CERT                          +49-711-685-5973/fax +49-711-685-5898

Re: lcs ethernet driver source

[Arjan van de Ven]

Ulrich Weigand <Ulrich.Weigand@de.ibm.com> wrote:
> Sorry, at this point we are not allowed to publish the source code of the
> lcs and qeth drivers (due to the use of confidential hardware interface
> specifications).  We make those modules available only in binary form
> on our developerWorks web site.

I actually believed that IBM took Open Source and the community seriously.
Not releasing key parts of the S/390 architecture port make me very 
disappointed in IBM given the very massive and public claims of support.

It makes all the fuss IBM is making about how great Linux is look, well, 
fake. I know that large parts of IBM actually do take Open Source seriously 
but it's regrettable that an influential part of the S/390 division doesn't.

> Well, you're right that without network connection, there's not much
> useful you can do.  However [.. snip ..] you can also use a CTC or IUCV
> network interface. Especially in a virtual machine, you can simply define
> a virtual CTC or IUCV connection and access your network via those
> devices.

The drivers you mention suffer a severe performance hit and are also, as I
understand them, nothing more than a sort of pipe to a proprietary driver on
the other side of the VM barrier; this is actually worse as the VM for S/390
is rather expensive. So I have the choice between not being able to
recompile the kernel while using the binary only module, and paying a lot of
money for a piece of software that isn't as fast as the direct drivers. 

This makes dwmw2@infradead.org's assertion -- that the S/390 code in the
current kernel is actually useless -- be a very valid point. Imagine Linux
for the x86 architecture had PPP as it's only open source network driver.
Wouldn't that limit its acceptance and utility ? IBM customers for S/390
might feel the same way; it will help IBM and the customers a lot if such
essential drivers were Open Source. Creative ideas by users might leverage
the S/390 platform into areas IBM didn't think of before; having all basic
parts open sourced is essential for that.

Now it's not all bad. I'm not trying to paint IBM as the corporate bad guy.
Most of IBM is doing a lot of great work for Linux and I hope that the small
part of IBM that seems to be blocking these drivers sees the light soon. If
there is anything I can do to help expedite that, just let me know.

Greetings,
   Arjan van de Ven

Re: lcs ethernet driver source

[John Alvord]

On Thu, 30 Aug 2001, David Woodhouse wrote:

> On Thu, 30 Aug 2001, Alan Cox wrote:
> 
> > > Sorry, at this point we are not allowed to publish the source code of the
> > > lcs and qeth drivers (due to the use of confidential hardware interface
> > > specifications).  We make those modules available only in binary form
> > > on our developerWorks web site.
> > 
> > Is there any plan to change this ? 
> 
> Erm, Linux on S/390 runs as a virtual machine, doesn't it? Does a lack of 
> network drivers not render it completely useless?

It can run as a virtual machine under VM. It can also run native directly
on the (LPAR) hardware.

john

Re: lcs ethernet driver source

[Ulrich Weigand]

Alan Cox wrote:

>> Sorry, at this point we are not allowed to publish the source code of
the
>> lcs and qeth drivers (due to the use of confidential hardware interface
>> specifications).  We make those modules available only in binary form
>> on our developerWorks web site.
>
>Is there any plan to change this ?

This is not something we (the Linux for S/390 development team) can decide;
it's up to the hardware groups that 'own' the LCS / QDIO specifications
whether they allow to make these public.  As I said, at this point, we
are not allowed to open the specs; while it is conceivable that this might
change in the future, I'm not aware of any specific plan.

(Please note that I'm in no way speaking for IBM here; that's just my
personal opinion.)


Mit freundlichen Gruessen / Best Regards

Ulrich Weigand

--
  Dr. Ulrich Weigand
  Linux for S/390 Design & Development
  IBM Deutschland Entwicklung GmbH, Schoenaicher Str. 220, 71032 Boeblingen
  Phone: +49-7031/16-3727   ---   Email: Ulrich.Weigand@de.ibm.com

Re: lcs ethernet driver source

[Ulrich Weigand]

David Woodhouse wrote:

>Erm, Linux on S/390 runs as a virtual machine, doesn't it? Does a lack of
>network drivers not render it completely useless?

Well, you're right that without network connection, there's not much
useful you can do.  However, LCS and QETH are not the only options;
you can also use a CTC or IUCV network interface (and those drivers *are*
open source and in the tree); there's also a third-party open source driver
to access CLAW network devices (not in the official tree, but e.g. part of
the SuSE kernel).

Especially in a virtual machine, you can simply define a virtual CTC or
IUCV
connection and access your network via those devices.

If you are running Linux native on the S/390 (not in a virtual machine)
and have *only* physical LCS or QETH devices, you must use the binary-only
modules; but this doesn't appear to be fundamentally different from other
devices where only binary drivers exist ...


Mit freundlichen Gruessen / Best Regards

Ulrich Weigand

--
  Dr. Ulrich Weigand
  Linux for S/390 Design & Development
  IBM Deutschland Entwicklung GmbH, Schoenaicher Str. 220, 71032 Boeblingen
  Phone: +49-7031/16-3727   ---   Email: Ulrich.Weigand@de.ibm.com

Re: lcs ethernet driver source

[David Woodhouse]

On Thu, 30 Aug 2001, Alan Cox wrote:

> > Sorry, at this point we are not allowed to publish the source code of the
> > lcs and qeth drivers (due to the use of confidential hardware interface
> > specifications).  We make those modules available only in binary form
> > on our developerWorks web site.
> 
> Is there any plan to change this ? 

Erm, Linux on S/390 runs as a virtual machine, doesn't it? Does a lack of 
network drivers not render it completely useless?

If the architecture port is not sanely usable in the form in which it 
appears in the Linux kernel, then it should be removed from the tree.

-- 
dwmw2

Re: lcs ethernet driver source

[Richard B. Johnson]

On Thu, 30 Aug 2001, Ulrich Weigand wrote:

> 
> Arjan van de Ven wrote:
> 
> >I tried to find the source for the lcs and qeth ethernet drivers for IBM's
> >S390 architecture, so I could recompile the kernel. However I seem to not
> be
> >able to find the source of these. Does anyone have a URL handy for this ?
> >(I wonder why these aren't in Linus or Alan's tree like most of the other
> >s390 patches)
> 
> Sorry, at this point we are not allowed to publish the source code of the
> lcs and qeth drivers (due to the use of confidential hardware interface
> specifications).  We make those modules available only in binary form
> on our developerWorks web site.
> 
Gosh. I didn't know you guys were so advanced that you didn't use
an electronic hardware interface! Your 'hardware interface specifications'
use magnetohydrodynamics, and they are top-secret, right?

Cheers,
Dick Johnson

Penguin : Linux version 2.4.1 on an i686 machine (799.53 BogoMips).

    I was going to compile a list of innovations that could be
    attributed to Microsoft. Once I realized that Ctrl-Alt-Del
    was handled in the BIOS, I found that there aren't any.

Re: lcs ethernet driver source

[Alan Cox]

> Sorry, at this point we are not allowed to publish the source code of the
> lcs and qeth drivers (due to the use of confidential hardware interface
> specifications).  We make those modules available only in binary form
> on our developerWorks web site.

Is there any plan to change this ? 

Re: lcs ethernet driver source

[Ulrich Weigand]

Arjan van de Ven wrote:

>I tried to find the source for the lcs and qeth ethernet drivers for IBM's
>S390 architecture, so I could recompile the kernel. However I seem to not
be
>able to find the source of these. Does anyone have a URL handy for this ?
>(I wonder why these aren't in Linus or Alan's tree like most of the other
>s390 patches)

Sorry, at this point we are not allowed to publish the source code of the
lcs and qeth drivers (due to the use of confidential hardware interface
specifications).  We make those modules available only in binary form
on our developerWorks web site.


Mit freundlichen Gruessen / Best Regards

Ulrich Weigand

--
  Dr. Ulrich Weigand
  Linux for S/390 Design & Development
  IBM Deutschland Entwicklung GmbH, Schoenaicher Str. 220, 71032 Boeblingen
  Phone: +49-7031/16-3727   ---   Email: Ulrich.Weigand@de.ibm.com