GPL License violation

[Jonathan Buzzard]

I received an email yesterday asking about the license of some code
found in FreeBSD to turn the fan on a Dell laptop on/off, the code in
question is shown below. I do not know exactly where in FreeBSD this
code lies, as I do not use FreeBSD. Neither do I want or should have to
go looking for it.

This code has been lifted verbatim out of drivers/char/i8k.c in Linux,
which in itself was a minor modification of code in
drivers/char/toshiba.c mdae by myself. This code dates back to January
1998, and was first included in Linux around 2.2.20. I am the original
author of this code, and it is clearly and only ever licensed under the
GPL. I must insist that it be removed forwith from FreeBSD.

JAB.


(cut along the line)
-------8<-------8<-------8<-------8<-------8<-------8<-------8<-------8<-------

        asm("pushl %%eax\n\t" \
       "movl 0(%%eax),%%edx\n\t" \
           "push %%edx\n\t" \
           "movl 4(%%eax),%%ebx\n\t" \
           "movl 8(%%eax),%%ecx\n\t" \
           "movl 12(%%eax),%%edx\n\t" \
           "movl 16(%%eax),%%esi\n\t" \
           "movl 20(%%eax),%%edi\n\t" \
           "popl %%eax\n\t" \
           "out %%al,$0xb2\n\t" \
           "out %%al,$0x84\n\t" \
           "xchgl %%eax,(%%esp)\n\t"
           "movl %%ebx,4(%%eax)\n\t" \
           "movl %%ecx,8(%%eax)\n\t" \
           "movl %%edx,12(%%eax)\n\t" \
           "movl %%esi,16(%%eax)\n\t" \
           "movl %%edi,20(%%eax)\n\t" \
           "popl %%edx\n\t" \
           "movl %%edx,0(%%eax)\n\t" \
           "lahf\n\t" \
           "shrl $8,%%eax\n\t" \
           "andl $1,%%eax\n" \
            : "=a" (rc)
            : "a" (regs)
            : "%ebx", "%ecx", "%edx", "%esi", "%edi", "memory");


    if ((rc != 0) || ((regs->eax & 0xffff) == 0xffff) || (regs->eax ==
eax))
    {
    return -1;
    }

-------8<-------8<-------8<-------8<-------8<-------8<-------8<-------8<-------



--
Jonathan A. Buzzard                 Email: jonathan (at) buzzard.me.uk
Northumberland, United Kingdom.       Tel: [phone]
_______________________________________________
freebsd-audit@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-audit
To unsubscribe, send any mail to "freebsd-audit-unsubscribe@..."

[Brooks Davis]

On Sun, Oct 08, 2006 at 01:29:25PM +0100, Jonathan Buzzard wrote:

> I received an email yesterday asking about the license of some code
> found in FreeBSD to turn the fan on a Dell laptop on/off, the code in
> question is shown below. I do not know exactly where in FreeBSD this
> code lies, as I do not use FreeBSD. Neither do I want or should have to
> go looking for it.
>
> This code has been lifted verbatim out of drivers/char/i8k.c in Linux,
> which in itself was a minor modification of code in
> drivers/char/toshiba.c mdae by myself. This code dates back to January
> 1998, and was first included in Linux around 2.2.20. I am the original
> author of this code, and it is clearly and only ever licensed under the
> GPL. I must insist that it be removed forwith from FreeBSD.
>
> JAB.

For the list.  Core has seen this and is investigating.

-- Brooks

[Lutz Böhne]

Dear Jonathan, -hackers and -chat

first of all, the freebsd-audit mailinglist has been dead for a few
months now (I wonder why I'm even still subscribed, I think it has been
officially retired some time ago) and your e-mail is not related to
"FreeBSD Security Auditing" either.

I'm not quite sure where your message would fit best, so for now I'm
CCing freebsd-hackers and freebsd-chat.

Regarding your issue: I tried several greps on a fresh copy of the
FreeBSD RELENG_6 source tree and couldn't find the code you're referring
to.

For instance, egrepping recursively for "movl.*20.*eax.*edi" (which
finds the line  "movl 20(%%eax),%%edi\n\t" in Linuxes i8k.c just fine)
will turn up any results.

Also, grepping recursively for just the "lahf" instruction wouldn't find
it in any context similar to the one you attached.

Grepping for Dell or Toshiba didn't turn up anything useful either.

There are at least three possibilities now:
- my grep skills are lacking, or
- the code is gone from the FreeBSD source (or it is not in RELENG_6,
but -CURRENT), or
- the person who sent you that e-mail is mistaken.

In any case, I suggest you get back to that person and ask for more
information.


Regards,

Lutz

Jonathan Buzzard wrote:

> I received an email yesterday asking about the license of some code
> found in FreeBSD to turn the fan on a Dell laptop on/off, the code in
> question is shown below. I do not know exactly where in FreeBSD this
> code lies, as I do not use FreeBSD. Neither do I want or should have to
> go looking for it.
>
> This code has been lifted verbatim out of drivers/char/i8k.c in Linux,
> which in itself was a minor modification of code in
> drivers/char/toshiba.c mdae by myself. This code dates back to January
> 1998, and was first included in Linux around 2.2.20. I am the original
> author of this code, and it is clearly and only ever licensed under the
> GPL. I must insist that it be removed forwith from FreeBSD.
>
> JAB.
>
>
> (cut along the line)
> -------8<-------8<-------8<-------8<-------8<-------8<-------8<-------8<-------
>
> asm("pushl %%eax\n\t" \
>        "movl 0(%%eax),%%edx\n\t" \
>   "push %%edx\n\t" \
>   "movl 4(%%eax),%%ebx\n\t" \
>   "movl 8(%%eax),%%ecx\n\t" \
>   "movl 12(%%eax),%%edx\n\t" \
>   "movl 16(%%eax),%%esi\n\t" \
>   "movl 20(%%eax),%%edi\n\t" \
>   "popl %%eax\n\t" \
>   "out %%al,$0xb2\n\t" \
>   "out %%al,$0x84\n\t" \
>   "xchgl %%eax,(%%esp)\n\t"
>   "movl %%ebx,4(%%eax)\n\t" \
>   "movl %%ecx,8(%%eax)\n\t" \
>   "movl %%edx,12(%%eax)\n\t" \
>   "movl %%esi,16(%%eax)\n\t" \
>   "movl %%edi,20(%%eax)\n\t" \
>   "popl %%edx\n\t" \
>   "movl %%edx,0(%%eax)\n\t" \
>   "lahf\n\t" \
>   "shrl $8,%%eax\n\t" \
>   "andl $1,%%eax\n" \
>    : "=a" (rc)
>    : "a" (regs)
>    : "%ebx", "%ecx", "%edx", "%esi", "%edi", "memory");
>
>
>     if ((rc != 0) || ((regs->eax & 0xffff) == 0xffff) || (regs->eax ==
> eax))
>     {
>     return -1;
>     }
>
> -------8<-------8<-------8<-------8<-------8<-------8<-------8<-------8<-------
>
>
>

--
Lutz Boehne - http://www.damogran.de

_______________________________________________
freebsd-audit@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-audit
To unsubscribe, send any mail to "freebsd-audit-unsubscribe@..."

[M. Warner Losh]

In message: <4528EF25.1000103@...>
            Jonathan Buzzard <jonathan@...> writes:
: I received an email yesterday asking about the license of some code

FYI: I've searched the FreeBSD tree for this code, and can't find it.
I've sent a longer message to jonathan asking for specific details...
Many folks have forwarded this to core@ and we'll discover what's
going on here...

Warner
_______________________________________________
freebsd-audit@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-audit
To unsubscribe, send any mail to "freebsd-audit-unsubscribe@..."

[M. Warner Losh]

In message: <45298188.1080201@...>
            Lutz Boehne <lboehne@...> writes:
: In any case, I suggest you get back to that person and ask for more
: information.

That was my suggestion as well.  I grepped the entire tree, both ports
and src, and couldn't find any code that came close to resembling this
code.

Googling shows two packages that contain this code, but those aren't
part of FreeBSD and they are distributed under the LGPL and the GPL.

Finally, there was no need to forward this to hackers@ or chat@.  Not
only did I see it in audit, several people forwarded this to core@ for
investigation.

Warner
_______________________________________________
freebsd-audit@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-audit
To unsubscribe, send any mail to "freebsd-audit-unsubscribe@..."

[Bruno Ducrot]

On Sun, Oct 08, 2006 at 05:01:30PM -0600, M. Warner Losh wrote:

> In message: <45298188.1080201@...>
>             Lutz Boehne <lboehne@...> writes:
> : In any case, I suggest you get back to that person and ask for more
> : information.
>
> That was my suggestion as well.  I grepped the entire tree, both ports
> and src, and couldn't find any code that came close to resembling this
> code.
>
> Googling shows two packages that contain this code, but those aren't
> part of FreeBSD and they are distributed under the LGPL and the GPL.
>
> Finally, there was no need to forward this to hackers@ or chat@.  Not
> only did I see it in audit, several people forwarded this to core@ for
> investigation.
>

I'm working on Dell's laptop support, even though I'm not
the one who code a tool for a fan control (and I don't
know if such tool under FreeBSD exist).

Some preminaly code can be found here:

http://people.FreeBSD.org/~bruno/i8kutils_bsd.tar.bz  [1]
http://people.FreeBSD.org/~bruno/acpi_dell.tar.gz     [2]
http://people.FreeBSD.org/~bruno/dellctl.tar.gz       [3]

For now, the 3 tar ball above have not been publically send to any
public list I'm aware of, because those are only priminally work.

For [1], people can check I haven't removed any copyright, nor I even
bothered adding my name.  In any case, I don't plan to add that one to
the base system.
In fact, I think to remove it from http://people.FreeBSD.org/~bruno/
in the near future.

For [2], people can check it's a really preliminary work, and is based
on some calls to ACPI methods under the DSDT.  Since it's a really
different approach taken from the driver found under Linux, it's free
from any GPL'ed code.

Finally [3] is only a userspace tool to control [2].

Since [2] and [3] are free from any GPL'ed codes, I consider commiting
them if one day they work.

Actually I even considered to port [2] under Linux, because this is
the right way to go when ACPI mode is enabled for obvious reason.
The io ports related to the SMM handler are shared, and ACPI take
care to handle an ACPI mutex before entering SMM, that at least might
eliminate strange cases when sometimes i8k doesn't work.

Cheers,

--
Bruno Ducrot

--  Which is worse:  ignorance or apathy?
--  Don't know.  Don't care.
_______________________________________________
freebsd-audit@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-audit
To unsubscribe, send any mail to "freebsd-audit-unsubscribe@..."

[Brett Glass]

Isn't it funny how quick the proponents of what the FSF calls
"Free" software (note the capital "F", which by itself is a sure
sign of propaganda or dogma) are to try to restrict its use?

--Brett Glass

_______________________________________________
freebsd-audit@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-audit
To unsubscribe, send any mail to "freebsd-audit-unsubscribe@..."

[Jonathan Buzzard]

Bruno Ducrot wrote:

[SNIP]

>
> I'm working on Dell's laptop support, even though I'm not
> the one who code a tool for a fan control (and I don't
> know if such tool under FreeBSD exist).
>
> Some preminaly code can be found here:
>
> http://people.FreeBSD.org/~bruno/i8kutils_bsd.tar.bz  [1]
> http://people.FreeBSD.org/~bruno/acpi_dell.tar.gz     [2]
> http://people.FreeBSD.org/~bruno/dellctl.tar.gz       [3]
>
> For now, the 3 tar ball above have not been publically send to any
> public list I'm aware of, because those are only priminally work.
>
> For [1], people can check I haven't removed any copyright, nor I even
> bothered adding my name.  In any case, I don't plan to add that one to
> the base system.
> In fact, I think to remove it from http://people.FreeBSD.org/~bruno/
> in the near future.
>
> For [2], people can check it's a really preliminary work, and is based
> on some calls to ACPI methods under the DSDT.  Since it's a really
> different approach taken from the driver found under Linux, it's free
> from any GPL'ed code.
>
> Finally [3] is only a userspace tool to control [2].
>
> Since [2] and [3] are free from any GPL'ed codes, I consider commiting
> them if one day they work.
>
> Actually I even considered to port [2] under Linux, because this is
> the right way to go when ACPI mode is enabled for obvious reason.
> The io ports related to the SMM handler are shared, and ACPI take
> care to handle an ACPI mutex before entering SMM, that at least might
> eliminate strange cases when sometimes i8k doesn't work.

There is already a Toshiba ACPI module that does that I believe. Or at
least it exposes a /dev/toshiba that enables you to but the laptop into
SMM in the same was as the Toshiba/Dell drivers do.

I still use a 2.4 kernel and hence APM as I have never gotten around to
upgrading to 2.6, as this is complicated by running an entirely LVM system.

You are however correct that ACPI is the correct way to go.

JAB.

--
Jonathan A. Buzzard                 Email: jonathan (at) buzzard.me.uk
Northumberland, United Kingdom.       Tel: [phone]
_______________________________________________
freebsd-audit@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-audit
To unsubscribe, send any mail to "freebsd-audit-unsubscribe@..."

[Ganbold]

Bruno Ducrot wrote:

> I'm working on Dell's laptop support, even though I'm not
> the one who code a tool for a fan control (and I don't
> know if such tool under FreeBSD exist).
>
> Some preminaly code can be found here:
>
> http://people.FreeBSD.org/~bruno/i8kutils_bsd.tar.bz  [1]
> http://people.FreeBSD.org/~bruno/acpi_dell.tar.gz     [2]
> http://people.FreeBSD.org/~bruno/dellctl.tar.gz       [3]
>
>  

Bruno,

Did you make suspend/resume work?
Did you make your volume up/down key work?
I have Latitude D620  and can't make above work.

thanks,

Ganbold

> For now, the 3 tar ball above have not been publically send to any
> public list I'm aware of, because those are only priminally work.
>
> For [1], people can check I haven't removed any copyright, nor I even
> bothered adding my name.  In any case, I don't plan to add that one to
> the base system.
> In fact, I think to remove it from http://people.FreeBSD.org/~bruno/
> in the near future.
>
> For [2], people can check it's a really preliminary work, and is based
> on some calls to ACPI methods under the DSDT.  Since it's a really
> different approach taken from the driver found under Linux, it's free
> from any GPL'ed code.
>
> Finally [3] is only a userspace tool to control [2].
>
> Since [2] and [3] are free from any GPL'ed codes, I consider commiting
> them if one day they work.
>
> Actually I even considered to port [2] under Linux, because this is
> the right way to go when ACPI mode is enabled for obvious reason.
> The io ports related to the SMM handler are shared, and ACPI take
> care to handle an ACPI mutex before entering SMM, that at least might
> eliminate strange cases when sometimes i8k doesn't work.
>
> Cheers,
>  

_______________________________________________
freebsd-audit@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-audit
To unsubscribe, send any mail to "freebsd-audit-unsubscribe@..."

[Bruno Ducrot]

On Mon, Oct 09, 2006 at 08:53:30PM +0100, Jonathan Buzzard wrote:

> Bruno Ducrot wrote:
>
> [SNIP]
> >
> > I'm working on Dell's laptop support, even though I'm not
> > the one who code a tool for a fan control (and I don't
> > know if such tool under FreeBSD exist).
> >
> > Some preminaly code can be found here:
> >
> > http://people.FreeBSD.org/~bruno/i8kutils_bsd.tar.bz  [1]
> > http://people.FreeBSD.org/~bruno/acpi_dell.tar.gz     [2]
> > http://people.FreeBSD.org/~bruno/dellctl.tar.gz       [3]
> >
> > For now, the 3 tar ball above have not been publically send to any
> > public list I'm aware of, because those are only priminally work.
> >
> > For [1], people can check I haven't removed any copyright, nor I even
> > bothered adding my name.  In any case, I don't plan to add that one to
> > the base system.
> > In fact, I think to remove it from http://people.FreeBSD.org/~bruno/
> > in the near future.
> >
> > For [2], people can check it's a really preliminary work, and is based
> > on some calls to ACPI methods under the DSDT.  Since it's a really
> > different approach taken from the driver found under Linux, it's free
> > from any GPL'ed code.
> >
> > Finally [3] is only a userspace tool to control [2].
> >
> > Since [2] and [3] are free from any GPL'ed codes, I consider commiting
> > them if one day they work.
> >
> > Actually I even considered to port [2] under Linux, because this is
> > the right way to go when ACPI mode is enabled for obvious reason.
> > The io ports related to the SMM handler are shared, and ACPI take
> > care to handle an ACPI mutex before entering SMM, that at least might
> > eliminate strange cases when sometimes i8k doesn't work.
>
> There is already a Toshiba ACPI module that does that I believe. Or at
> least it exposes a /dev/toshiba that enables you to but the laptop into
> SMM in the same was as the Toshiba/Dell drivers do.

Not exactly.  The ACPI method for the Toshiba is not the same as for the
Dell.

Toshiba use \_SB_.VALD.GHCI() method in order to access the HCI whereas
Dell use \SMI() (and other methods).  They are different and don't share
any API I'm aware of, but I may be wrong of course.

> I still use a 2.4 kernel and hence APM as I have never gotten around to
> upgrading to 2.6, as this is complicated by running an entirely LVM system.
>
> You are however correct that ACPI is the correct way to go.

Depend.  If APM work better for you there is no reason to enable ACPI
after all.

Cheers,

--
Bruno Ducrot

--  Which is worse:  ignorance or apathy?
--  Don't know.  Don't care.
_______________________________________________
freebsd-audit@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-audit
To unsubscribe, send any mail to "freebsd-audit-unsubscribe@..."

[Bruno Ducrot]

On Tue, Oct 10, 2006 at 02:45:21PM +0800, Ganbold wrote:
> Bruno,
>
> Did you make suspend/resume work?

Unrelated.  acpi_dell isn't for suspend/resume.
But I'm aware of the problem and I'm working on that.

> Did you make your volume up/down key work?

That one of the goal.  But there is a need to work more and I'm waiting
some input from someone in touch with Dell BIOS engineers.
That why I said this doesn't work, and why I haven't yet
made them public.
But I don't know if Jonathan Buzzard was aware of that work
and if his complain was related to this work.

> I have Latitude D620  and can't make above work.

I might get one D620 soon hopefully.  I will then be able to work more
on this support.

--
Bruno Ducrot

--  Which is worse:  ignorance or apathy?
--  Don't know.  Don't care.
_______________________________________________
freebsd-audit@... mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-audit
To unsubscribe, send any mail to "freebsd-audit-unsubscribe@..."