Re: Suggestion for handling of boot-time driver options

Stephen Williams (steve@icarus.icarus.com)
Sun, 14 Dec 1997 16:13:40 -0800


> They *could* be written entirely from scratch, and be quite
> independend of actual kernel. (It would be pretty hard to write kernel
> module that way, as they would have to re-implement all includes, but
> possible). Where do they violate gpl, then?

alan@lxorguk.ukuu.org.uk said:
> By being linked with it

----- From COPYING, section 2:

If identifiable sections of that work are not derived from the Program,
and can be reasonably considered independent and separate works in
themselves, then this License, and its terms, do not apply to those
sections when you distribute them as separate works.
-----

Aftermarket device drivers are (presumably) distributed separately so I
would think that this clause applies. Even if the .o of the binary-only
driver is included on a disk that also has Linux complete with source,

"... mere aggregation of another work not based on the Program
with the Program [...] on a volume of a storage or distribution
medium does not bring the other work under the scope of this
License."

So does loading a module into a running kernel make the now running system
a derivative work? What is the difference between linking at run time and
bolting it on by ELF magic before booting?

Anyhow, I would not advocate that someone be tight-lipped about device driver
source code. It *would* be nice to be able to distribute compiled driver
modules like one can now distribute compiled emacs, or compiled gcc.

(To be clear, Picture Elements applies GPL to device drivers for boards
we produce. We believe it is the right thing to do, even though we do not
believe we are obliged to do so. Well, technologically we are because pre-
compiled modules cannot be transported practically.)

-- 
Steve Williams
steve@icarus.com
steve@picturel.com

"The woods are lovely, dark and deep. But I have promises to keep, And lines to code before I sleep, And lines to code before I sleep."