Re: [PATCH 2.5.5] do export vmalloc_to_page to modules...

[Ingo Molnar <mingo@redhat.com>]

On Thu, 4 Apr 2002, Andrea Arcangeli wrote:

> adding export symbol here and there it's the same thing you did in the
> redhat kernel and in your tux patches here:

it was done by first *asking* all maintainers/authors involved, including
the network folks and Linus. Plus at the time it was done no _GPL way to
signal internal components existed, i'd otherwise have used it to document
that this is an internal export only, for the fully GPL-ed TUX subsystem.
But i'd have no problem with making TUX a fully statically linked thing
either. [it's just so convenient to demand-link TUX as a part of the
kernel.]

> There is no difference at all with what you did above and with my
> removal of the _GPL tag from the vmalloc_to_page [...]

(lets stop this vmalloc_to_page() thing, as i said i agree with providing
it as a published export. It's a small-enough function to be a non-issue.)

> Now if my understanding is wrong, I'd like to know of course, I'm not
> expert here, but the only logical thing I'm sure about is that if it's
> illegal for me to export my GPL wrapper then I've just the right to make
> all non GPL drivers illegal, that is the only logical sure thing that
> can be deducted. And yes, I'd be really happy if I'd that right.

while i'm not a lawyer either, i think the question here is intent, like
in most matters of law/contract. If your intent is to make something that
is a derivative look like something that is not a derivative, you are on
the bad side. Eg. the GPL also uses intent in a form - eg. 'source code'
is not some arbitrary language or format, 'source code' is the preferred
form intended for development. In this sense it's not a GPL-conform
publication of source code to provide hex-encoded objects within C files,
even though C code matches the technical definition of 'source code'.

the other problem is that i think we really want to cooperate with people
who'd like to interface with the kernel in kernel-space, without making
their code a derivative of the kernel, along a well-defined API, even if
those people do not want to GPL their code for whatever reasons. But like
Alex mentioned it, Linux never had a 'well defined module API'. There was
no guarantee, no nothing, it's not an API in the GPL sense i think.

	Ingo