Re: LGPL v3 compatibilty

[Walter Landry <wlandry@ucsd.edu>]

Måns_Rullgård <mans@mansr.com> wrote:
> Walter Landry <walter@geodynamics.org> writes:
> 
> > Francesco Poli <frx@firenze.linux.it> wrote:
> >> On Sat, 14 Jul 2007 21:56:27 -0700 (PDT) Walter Landry wrote:
> >> 
> >> > Francesco Poli <frx@firenze.linux.it> wrote:
> >> > > On Mon, 2 Jul 2007 12:31:13 -0400 Anthony Towns wrote:
> >> > > 
> >> > > [...]
> >> > > > Note that _if_ we do stick to the view we've taken up until now,
> >> > > > when we have a LGPLv3 only glibc in the archive, we'll no longer
> >> > > > be able to distribute GPLv2-only compiled executables.
> >> > > 
> >> > > Unless the GPLv2-only work copyright holder(s) add(s) a special
> >> > > exception, similar to the one needed to link with the OpenSSL
> >> > > library, right?
> >> > > 
> >> > > This scenario is worrying me...  :-(
> >> > 
> >> > Is this going to be a problem for the kernel?  It is definitely not
> >> > going to go to GPLv3.
> >> 
> >> Is the Linux kernel linked with any LGPL'd work?
> >> AFAIUI, it is not, so no problem for the kernel.
> >
> > Doesn't the kernel get its implementations for pow(), sqrt(),
> > printf(), and the rest of the C standard library from glibc, which is
> > LGPL'd?
> 
> No.  The kernel is completely self-contained.  Some code may of course
> have been borrowed from glibc at some point, but that's irrelevant.

Are you sure that it is self-contained?  Grepping through the sources
of 2.6.22.1, I do not see an implementation of <stdarg.h> or
<stdio.h>.  I do see <string.h>, and <math.h> is never included.

Cheers,
Walter Landry
wlandry@ucsd.edu