This patch adds the missing MODULE_LICENSE("GPL").
Reported-by: Adrian Bunk <[email protected]>
Signed-off-by: Adrian Bunk <[email protected]>
---
This patch has been sent on:
- 15 Jun 2008
4cec562181f818f16f072b9a14ca7a0ce6c66417 diff --git a/fs/binfmt_som.c b/fs/binfmt_som.c
index fdc36bf..6d3bebb 100644
--- a/fs/binfmt_som.c
+++ b/fs/binfmt_som.c
@@ -308,3 +308,5 @@ static void __exit exit_som_binfmt(void)
core_initcall(init_som_binfmt);
module_exit(exit_som_binfmt);
+
+MODULE_LICENSE("GPL");
[email protected]
On Thu, Aug 07, 2008 at 12:11:55AM +0300, Adrian Bunk wrote:
> This patch adds the missing MODULE_LICENSE("GPL").
>
> Reported-by: Adrian Bunk <[email protected]>
> Signed-off-by: Adrian Bunk <[email protected]>
>
> ---
>
> This patch has been sent on:
> - 15 Jun 2008
>
> 4cec562181f818f16f072b9a14ca7a0ce6c66417 diff --git a/fs/binfmt_som.c b/fs/binfmt_som.c
> index fdc36bf..6d3bebb 100644
> --- a/fs/binfmt_som.c
> +++ b/fs/binfmt_som.c
> @@ -308,3 +308,5 @@ static void __exit exit_som_binfmt(void)
>
> core_initcall(init_som_binfmt);
> module_exit(exit_som_binfmt);
> +
> +MODULE_LICENSE("GPL");
Someone from HP needs to ACK this patch since I believe HP funded the code
changes made during the parisc-linux port. This should be a no-brainer
and I expect Bdale Garbee can quickly take care of this.
thanks,
grant
On Sat, Aug 09, 2008 at 06:55:11PM -0600, Grant Grundler wrote:
> Someone from HP needs to ACK this patch since I believe HP funded the code
> changes made during the parisc-linux port. This should be a no-brainer
> and I expect Bdale Garbee can quickly take care of this.
I actually did the work here, it was derived heavily from binfmt_elf
and binfmt_aout; had it not been GPL-licensed, it would not have been
distributable. Also, I believe I did this work before joining Linuxcare,
so I don't believe HP has any claim to it at all. You'd have to check
CVS history to be sure. Oh, and even if HP does have a claim to it
through Linuxcare, HP's contract required all changes made by Linuxcare
to be contributed upstream. So three different reasons why this code
being GPL is required.
--
Intel are signing my paycheques ... these opinions are still mine
"Bill, look, we understand that you're interested in selling us this
operating system, but compare it to ours. We can't possibly take such
a retrograde step."
On Sat, 2008-08-09 at 19:00 -0600, Matthew Wilcox wrote:
> On Sat, Aug 09, 2008 at 06:55:11PM -0600, Grant Grundler wrote:
> > Someone from HP needs to ACK this patch since I believe HP funded the code
> > changes made during the parisc-linux port. This should be a no-brainer
> > and I expect Bdale Garbee can quickly take care of this.
> I believe I did this work before joining Linuxcare,
> so I don't believe HP has any claim to it at all. You'd have to check
> CVS history to be sure.
If that is indeed the timing, I agree. If not, let me know, and I'll
see if I can push the right buttons in HP.
Bdale
On Sat, Aug 09, 2008 at 06:55:11PM -0600, Grant Grundler wrote:
> [email protected]
>
> On Thu, Aug 07, 2008 at 12:11:55AM +0300, Adrian Bunk wrote:
> > This patch adds the missing MODULE_LICENSE("GPL").
> >
> > Reported-by: Adrian Bunk <[email protected]>
> > Signed-off-by: Adrian Bunk <[email protected]>
> >
> > ---
> >
> > This patch has been sent on:
> > - 15 Jun 2008
> >
> > 4cec562181f818f16f072b9a14ca7a0ce6c66417 diff --git a/fs/binfmt_som.c b/fs/binfmt_som.c
> > index fdc36bf..6d3bebb 100644
> > --- a/fs/binfmt_som.c
> > +++ b/fs/binfmt_som.c
> > @@ -308,3 +308,5 @@ static void __exit exit_som_binfmt(void)
> >
> > core_initcall(init_som_binfmt);
> > module_exit(exit_som_binfmt);
> > +
> > +MODULE_LICENSE("GPL");
>
> Someone from HP needs to ACK this patch since I believe HP funded the code
> changes made during the parisc-linux port. This should be a no-brainer
> and I expect Bdale Garbee can quickly take care of this.
Either the file is available since the beginning under the terms of the
GPLv2 (or a compatible license that can be reduced to GPLv2) or it's
addition to the kernel was a huge mistake.
> thanks,
> grant
cu
Adrian
--
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon Seed
On Sun, Aug 10, 2008 at 10:41:54AM -0300, Bdale Garbee wrote:
> On Sat, 2008-08-09 at 19:00 -0600, Matthew Wilcox wrote:
> > On Sat, Aug 09, 2008 at 06:55:11PM -0600, Grant Grundler wrote:
> > > Someone from HP needs to ACK this patch since I believe HP funded the code
> > > changes made during the parisc-linux port. This should be a no-brainer
> > > and I expect Bdale Garbee can quickly take care of this.
>
> > I believe I did this work before joining Linuxcare,
> > so I don't believe HP has any claim to it at all. You'd have to check
> > CVS history to be sure.
>
> If that is indeed the timing, I agree. If not, let me know, and I'll
> see if I can push the right buttons in HP.
http://cvs.parisc-linux.org/obsolete/linux-2.2/fs/binfmt_som.c?rev=1.8&view=log
indicates I first committed the file in June 1999 and was the majority
committer to this file (prumpf also committed some code). My last
commit to it was in November 1999 (while I still worked for Genedata).
At some point around then, we moved to the 2.4 CVS tree:
http://cvs.parisc-linux.org/linux-2.4/fs/binfmt_som.c?rev=1.21&view=log
and there are some commits from other people at HP and myself while at
Linuxcare and HP.
The 2.5 development tree:
http://cvs.parisc-linux.org/linux-2.5/fs/binfmt_som.c?rev=1.3&view=log
doesn't show anything of great interest.
and the 2.6 tree represents accurately how much effort we put into SOM:
http://cvs.parisc-linux.org/linux-2.6/fs/binfmt_som.c?rev=1.8&view=log
So while HP has a legitimate claim to some parts of the file, my initial
contributions before I worked for Linuxcare or HP were the largest and I
don't think that anyone has a legitimate claim that this is not a
derived work of GPL code, and hence I think Adrian's patch should be
accepted.
--
Intel are signing my paycheques ... these opinions are still mine
"Bill, look, we understand that you're interested in selling us this
operating system, but compare it to ours. We can't possibly take such
a retrograde step."
On Sun, Aug 10, 2008 at 09:14:33AM -0600, Matthew Wilcox wrote:
...
> > If that is indeed the timing, I agree. If not, let me know, and I'll
> > see if I can push the right buttons in HP.
>
> http://cvs.parisc-linux.org/obsolete/linux-2.2/fs/binfmt_som.c?rev=1.8&view=log
>
> indicates I first committed the file in June 1999 and was the majority
> committer to this file (prumpf also committed some code). My last
> commit to it was in November 1999 (while I still worked for Genedata).
Ok - I didn't realize that.
> At some point around then, we moved to the 2.4 CVS tree:
>
> http://cvs.parisc-linux.org/linux-2.4/fs/binfmt_som.c?rev=1.21&view=log
>
> and there are some commits from other people at HP and myself while at
> Linuxcare and HP.
Thanks for digging this up.
I expect it should be easy for Bdale to get this rubber stamped
by any one of the HP linux kernel engineers. HP legal shouldn't
need to be involved.
>
> The 2.5 development tree:
> http://cvs.parisc-linux.org/linux-2.5/fs/binfmt_som.c?rev=1.3&view=log
>
> doesn't show anything of great interest.
>
> and the 2.6 tree represents accurately how much effort we put into SOM:
> http://cvs.parisc-linux.org/linux-2.6/fs/binfmt_som.c?rev=1.8&view=log
>
>
> So while HP has a legitimate claim to some parts of the file, my initial
> contributions before I worked for Linuxcare or HP were the largest and I
> don't think that anyone has a legitimate claim that this is not a
> derived work of GPL code, and hence I think Adrian's patch should be
> accepted.
I agree and apologize for my bad memory and lazy attitude. I didn't mean
to make a big deal of this - just assumed HP was the right party to
address this to.
cheers,
grant
>
> --
> Intel are signing my paycheques ... these opinions are still mine
> "Bill, look, we understand that you're interested in selling us this
> operating system, but compare it to ours. We can't possibly take such
> a retrograde step."
On Sun, Aug 10, 2008 at 05:11:59PM +0300, Adrian Bunk wrote:
> On Sat, Aug 09, 2008 at 06:55:11PM -0600, Grant Grundler wrote:
...
> > Someone from HP needs to ACK this patch since I believe HP funded the code
> > changes made during the parisc-linux port. This should be a no-brainer
> > and I expect Bdale Garbee can quickly take care of this.
>
> Either the file is available since the beginning under the terms of the
> GPLv2 (or a compatible license that can be reduced to GPLv2) or it's
> addition to the kernel was a huge mistake.
As a former HP employee who was one of the several parisc-linux developers,
I'm certain the intent was to release this as GPL. That's why I said
"this should be a no-brainer".
However, it's still HP's (and not my) right to assert that.
thanks,
grant
On Sun, Aug 10, 2008 at 06:10:15PM -0600, Grant Grundler wrote:
> On Sun, Aug 10, 2008 at 05:11:59PM +0300, Adrian Bunk wrote:
> > On Sat, Aug 09, 2008 at 06:55:11PM -0600, Grant Grundler wrote:
> ...
> > > Someone from HP needs to ACK this patch since I believe HP funded the code
> > > changes made during the parisc-linux port. This should be a no-brainer
> > > and I expect Bdale Garbee can quickly take care of this.
> >
> > Either the file is available since the beginning under the terms of the
> > GPLv2 (or a compatible license that can be reduced to GPLv2) or it's
> > addition to the kernel was a huge mistake.
>
> As a former HP employee who was one of the several parisc-linux developers,
> I'm certain the intent was to release this as GPL. That's why I said
> "this should be a no-brainer".
>
> However, it's still HP's (and not my) right to assert that.
If it wasn't already 100% clear that the file was released under
the terms of the GPLv2 (or a compatible license that can be reduced
to GPLv2) at the day when it was included into the kernel something
went horribly wrong back then.
Otherwise HP would have the situation SCO was desperately searching for...
> thanks,
> grant
cu
Adrian
--
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon Seed
> On Sun, Aug 10, 2008 at 06:10:15PM -0600, Grant Grundler wrote:
> > On Sun, Aug 10, 2008 at 05:11:59PM +0300, Adrian Bunk wrote:
> > > On Sat, Aug 09, 2008 at 06:55:11PM -0600, Grant Grundler wrote:
> > ...
> > > > Someone from HP needs to ACK this patch since I believe HP funded the code
> > > > changes made during the parisc-linux port. This should be a no-brainer
> > > > and I expect Bdale Garbee can quickly take care of this.
> > >
> > > Either the file is available since the beginning under the terms of the
> > > GPLv2 (or a compatible license that can be reduced to GPLv2) or it's
> > > addition to the kernel was a huge mistake.
> >
> > As a former HP employee who was one of the several parisc-linux developers,
> > I'm certain the intent was to release this as GPL. That's why I said
> > "this should be a no-brainer".
> >
> > However, it's still HP's (and not my) right to assert that.
>
> If it wasn't already 100% clear that the file was released under
> the terms of the GPLv2 (or a compatible license that can be reduced
> to GPLv2) at the day when it was included into the kernel something
> went horribly wrong back then.
>
> Otherwise HP would have the situation SCO was desperately searching for...
This file was included before the SCO situation arose. The copyright
notice is similar to most. Checking 2.6.22.19, I see that more than
half the .c files lack MODULE_LICENSE("GPL") lines.
I don't understand why we are talking about HP here. According to
Matthew, his last commit to this file occured while he worked for
Genedata. It is my understanding that copyright normally acrues
to employers in employment situations. If Matthew was an employee
and not an independent contractor at the time, then why are we not
talking about Genedata, or its subsequent owner? Subsequent
employment at HP or their funding of the parisc port shouldn't
affect the licensing of a file that was previously contributed.
The file appears to be a derived work. This may also affect its
copyright status. The only reason HP should be involved is if the
file somehow contains material copyrighted by HP.
The whole licensing issue for the linux program and modules,
particularly for individual files, appears to be a mess. The
COPYING file is vague on the licensing for files. The recommended
wording suggested by the FSF isn't used. I can't see that adding
MODULE_LICENSE("GPL") lines is going to help much.
There are several GPL licenses and versions. The line isn't
going to mean much to a lawyer. I really think each file should
be specific about its licensing in words that can be clearly
understood.
Dave
--
J. David Anglin [email protected]
National Research Council of Canada (613) 990-0752 (FAX: 952-6602)
On Sun, 2008-08-10 at 09:14 -0600, Matthew Wilcox wrote:
> So while HP has a legitimate claim to some parts of the file, my initial
> contributions before I worked for Linuxcare or HP were the largest and I
> don't think that anyone has a legitimate claim that this is not a
> derived work of GPL code, and hence I think Adrian's patch should be
> accepted.
I agree. Thanks for tracking down the details.
Bdale
On Mon, Aug 11, 2008 at 10:56:51AM -0400, John David Anglin wrote:
> > On Sun, Aug 10, 2008 at 06:10:15PM -0600, Grant Grundler wrote:
> > > On Sun, Aug 10, 2008 at 05:11:59PM +0300, Adrian Bunk wrote:
> > > > On Sat, Aug 09, 2008 at 06:55:11PM -0600, Grant Grundler wrote:
> > > ...
> > > > > Someone from HP needs to ACK this patch since I believe HP funded the code
> > > > > changes made during the parisc-linux port. This should be a no-brainer
> > > > > and I expect Bdale Garbee can quickly take care of this.
> > > >
> > > > Either the file is available since the beginning under the terms of the
> > > > GPLv2 (or a compatible license that can be reduced to GPLv2) or it's
> > > > addition to the kernel was a huge mistake.
> > >
> > > As a former HP employee who was one of the several parisc-linux developers,
> > > I'm certain the intent was to release this as GPL. That's why I said
> > > "this should be a no-brainer".
> > >
> > > However, it's still HP's (and not my) right to assert that.
> >
> > If it wasn't already 100% clear that the file was released under
> > the terms of the GPLv2 (or a compatible license that can be reduced
> > to GPLv2) at the day when it was included into the kernel something
> > went horribly wrong back then.
> >
> > Otherwise HP would have the situation SCO was desperately searching for...
>
> This file was included before the SCO situation arose. The copyright
> notice is similar to most. Checking 2.6.22.19, I see that more than
> half the .c files lack MODULE_LICENSE("GPL") lines.
MODULE_LICENSE() has nothing to do with the SCO case.
As the name already indicates it is only for modules (it doesn't make
sense in built-in code) for showing that a module has a licence that is
compatible with the GPL (all other modules are legally somewhere between
fishy and illegal).
>...
> The whole licensing issue for the linux program and modules,
> particularly for individual files, appears to be a mess. The
> COPYING file is vague on the licensing for files. The recommended
> wording suggested by the FSF isn't used. I can't see that adding
> MODULE_LICENSE("GPL") lines is going to help much.
>
> There are several GPL licenses and versions. The line isn't
> going to mean much to a lawyer. I really think each file should
> be specific about its licensing in words that can be clearly
> understood.
The toplevel COPYING file of the kernel sources contains the exact
licence text for all files (including binfmt_som.c) in the kernel.
Even if the wording suggested by the FSF is not used in all files it's
clear that the kernel is under the GPLv2, and there isn't any serious
doubt that if a file like binfmt_som.c was not under a GPLv2 compatible
licence but built into a GPLv2'ed kernel the resulting image couldn't be
legally distributed.
Does anyone involved with code going from HP to the Linux kernel
seriously think that there is code for which it isn't 100% certain
that it was made available under the GPLv2?
> Dave
cu
Adrian
--
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon Seed
> The toplevel COPYING file of the kernel sources contains the exact
> licence text for all files (including binfmt_som.c) in the kernel.
The COPYING file contains the license text but it doesn't say clearly
what the license applies to. The closest thing to this is the statement
Also note that the GPL below is copyrighted by the Free Software
Foundation, but the instance of code that it refers to (the Linux
kernel) is copyrighted by me and others who actually wrote it.
in the COPYING file. The applicability of the GPL is parenthetical.
The HOWTO file is more specific, but a clear license statement at the
top level of the "source" tree is missing.
Dave
--
J. David Anglin [email protected]
National Research Council of Canada (613) 990-0752 (FAX: 952-6602)
On Mon, Aug 11, 2008 at 10:56:51AM -0400, John David Anglin wrote:
> This file was included before the SCO situation arose. The copyright
> notice is similar to most. Checking 2.6.22.19, I see that more than
> half the .c files lack MODULE_LICENSE("GPL") lines.
>
> I don't understand why we are talking about HP here. According to
> Matthew, his last commit to this file occured while he worked for
> Genedata.
Whoops, that was in the linux-2.2 CVS repository. There were subsequent
commits in the 2.4/2.5/2.6 repositories, some of which HP might have
a claim to.
> It is my understanding that copyright normally acrues
> to employers in employment situations. If Matthew was an employee
> and not an independent contractor at the time, then why are we not
> talking about Genedata, or its subsequent owner?
While I don't have a copy of my employment contract with Genedata any
more, my recollection is that they did not claim copyright on works I
produced that were off company time. It would be governed by Swiss law
which I believe would not permit such a term in the contract.
> Subsequent
> employment at HP or their funding of the parisc port shouldn't
> affect the licensing of a file that was previously contributed.
That's my basic argument. If binfmt_elf had contained the GPL
boilerplate, I would have copied that when creating binfmt_som. It was
intended to be GPL, it was a derivative work of GPL file (that also
didn't explicitly state it was GPL), it has been treated as GPL for
almost a decade.
--
Intel are signing my paycheques ... these opinions are still mine
"Bill, look, we understand that you're interested in selling us this
operating system, but compare it to ours. We can't possibly take such
a retrograde step."
On Mon, Aug 11, 2008 at 12:56:43PM -0400, John David Anglin wrote:
> > The toplevel COPYING file of the kernel sources contains the exact
> > licence text for all files (including binfmt_som.c) in the kernel.
>
> The COPYING file contains the license text but it doesn't say clearly
> what the license applies to. The closest thing to this is the statement
>
> Also note that the GPL below is copyrighted by the Free Software
> Foundation, but the instance of code that it refers to (the Linux
> kernel) is copyrighted by me and others who actually wrote it.
>
> in the COPYING file. The applicability of the GPL is parenthetical.
> The HOWTO file is more specific, but a clear license statement at the
> top level of the "source" tree is missing.
There's a file named COPYING at the top level, and a file named README
at the top level that says:
WHAT IS LINUX?
...
It is distributed under the GNU General Public License - see the
accompanying COPYING file for more details.
The GPLv2 licence on the Linux kernel is generally agreed upon
(whether one likes it or not is a different question, but all
contributors are aware that it applies) and has even been successfully
enforced at court, which makes your attempts of finding a way to
possibly interpret it differently kinda strange.
> Dave
cu
Adrian
--
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon Seed
On Mon, Aug 11, 2008 at 10:56:51AM -0400, John David Anglin wrote:
...
> I don't understand why we are talking about HP here.
My bad. I thought HP owned the copy right to the bulk of the code
and Matthew demonstrated they do not.
> According to
> Matthew, his last commit to this file occured while he worked for
> Genedata. It is my understanding that copyright normally acrues
> to employers in employment situations. If Matthew was an employee
> and not an independent contractor at the time, then why are we not
> talking about Genedata, or its subsequent owner? Subsequent
> employment at HP or their funding of the parisc port shouldn't
> affect the licensing of a file that was previously contributed.
Correct. I thought the work was done by HP employees and that was wrong.
sorry,
grant
> The file appears to be a derived work. This may also affect its
> copyright status. The only reason HP should be involved is if the
> file somehow contains material copyrighted by HP.
>
> The whole licensing issue for the linux program and modules,
> particularly for individual files, appears to be a mess. The
> COPYING file is vague on the licensing for files. The recommended
> wording suggested by the FSF isn't used. I can't see that adding
> MODULE_LICENSE("GPL") lines is going to help much.
>
> There are several GPL licenses and versions. The line isn't
> going to mean much to a lawyer. I really think each file should
> be specific about its licensing in words that can be clearly
> understood.
>
> Dave
> --
> J. David Anglin [email protected]
> National Research Council of Canada (613) 990-0752 (FAX: 952-6602)