Copyright infringement in linux/drivers/usb/serial/keyspan*fw.h
This message sparked a long thread on the debian-legal mailing list, which is long since dead. I am personally very curious about whether this has been resolved upstream. I consider it a very important issue, which is why I asked for RMS' opinion. He said that what is being done is clearly not "mere aggregation", and that such firmware should be moved out of the kernel (and even the tarball) to stop violating the GPL and make Linux be free software.
As a user of hardware which requires firmware like this, I have mixed feelings, but feel strongly that requirements of the GPL clearly override any measure of convenience. Are there any plans to remove the binary-only firmware from the kernel, and/or eventually from the Linux source distribution? I am not refering to this USB driver specifically, but rather the general occurance of firmware embedded in Linux device trivers.
----- Forwarded message from "Adam J. Richter" <[email protected]> -----
From: "Adam J. Richter" <[email protected]>
Date: Sun, 22 Apr 2001 12:53:48 -0700
To: [email protected]
Subject: Copyright infringement in linux/drivers/usb/serial/keyspan*fw.h
Linus's Linux kernel releases from 2.3.50 through the latest
test release (2.4.4-pre6) contain GPL-incompatible "firmware" images for "EZUSB" devices in linux/drivers/usb/serial/keyspan*fw.h, which are #included into drivers which contain GPL'ed code, even when compiled as modules.
I believe this infringinges the copyrights of the authors
of the code used in these drivers who released their code under GPL. Alan Cox, has gone on a campaign claiming that this is "mere aggregation" and insists that I bring in the lawyers to prove him otherwise. I really do not want to take this step, but he is forcing my hand. Note that Yggdrasil is a copyright owner in this case.
To simplify removal of the offending code, I have provided
a separate user level facility that can use the USB "hot plugging" system to automatically load that "firmware" or any other. The USB serial maintainer already plans to switch to it in 2.5, and has tested it and verified that it works. The software and a kernel patch to remove the offending code is FTPable from ftp://ftp.yggdrasil.com/pub/dist/device_control/ezusb/. The kernel patch also moves the "whiteheat" code loading into the same user space utility, just for technical reasons, even though that code can apparently be legally included in the kernel. Note that even without this software, the EZUSB firmware can apparently be loaded by other facilities or from other operating systems.
Here is what people involved in Debian should do:
1. Make sure that no Debian release or snapshot that includes a kernel from 2.3.50 through the current one (don't know if there are any yet) includes linux/drivers/usb/serial/keyspan*fw.h. By the way, even if there were no legal liability, these files could not be in the "free" part of Debian, if I read the Debian Free Software "Guidelines" correctly.
2. Possibly include the user level software FTPable from Yggdrasil, although be warned that that code will probably be replaced in the near future to use an input file format more compatible with other development tools used for EZUSB 8051 microcontroller software development.
Adam J. Richter __ ______________ 4880 Stevens Creek Blvd, Suite 104 [email protected] \ / San Jose, California 95129-1034 +1 408 261-6630 | g g d r a s i l United States of America
fax +1 408 261-6631 "Free Software For The Rest Of Us."
Signature · 15 lines
To UNSUBSCRIBE, email to [email protected] with a subject of "unsubscribe". Trouble? Contact [email protected] ----- End forwarded message ----- -- #define m(i)(x[i]^s[i+84])<< /* I do not condone improper use of this code */ unsigned char x[5],y,s[2048];main(n){for(read(0,x,5);read(0,s,n=2048);write(1,s ,n))if(s[y=s[13]%8+20]/16%4==1){int i=m(1)17^256+m(0)8,k=m(2)0,j=m(4)17^m(3)9^k *2-k%8^8,a=0,c=26;for(s[y]-=16;--c;j*=2)a=a*2^i&1,i=i/2^j&1<<24;for(j=127;++j<n ;c=c>y)c+=y=i^i/8^i>>4^i>>12,i=i>>8^y<<17,a^=a>>14,y=a^a*8^a<<6,a=a>>8^y<<9,k=s [j],k="7Wo~'G_\216"[k&7]+2^"cr3sfw6v;*k+>/n."[k>>4]*2^k*257/8,s[j]=k^(k&k*2&34) *6^c+~y;}}//Please join us in civil disobedience and distribute DeCSS(or efdtt!)
Copyright for drivers- two SCSI HBA drivers
Quoted 9 lines
As a user of hardware which requires firmware like this, I have mixed feelings, but feel strongly that requirements of the GPL clearly override any measure of convenience. Are there any plans to remove the binary-only firmware from the kernel, and/or eventually from the Linux source distribution? I am not refering to this USB driver specifically, but rather the general occurance of firmware embedded in Linux device trivers.
In the specific instances that I know about, the f/w for the older Qlogic SCSI cards (the drivers *not* supplied by vendors) for qlogicisp (qlogicisp_asm.c) and qlogicpti (qlogicisp_pti.c) is in the Linux source *and has been for years) with no copyright attribution whatsoever. qlogicfc has the BSD style copyright that I partly, but mostly Theo Deraadt of OpenBSD, managed to beat QLogic into doing.
Versions of f/w for qlogicisp && qlogicpti can easily also be had with the BSD licence- check any *BSD distribution, or pick 'em up via bitkeeper from my site.
-matt
Quoted 7 lines · Aaron Lehmann
This message sparked a long thread on the debian-legal mailing list, which is long since dead. I am personally very curious about whether this has been resolved upstream. I consider it a very important issue, which is why I asked for RMS' opinion. He said that what is being done is clearly not "mere aggregation", and that such firmware should be moved out of the kernel (and even the tarball) to stop violating the GPL and make Linux be free software.
Last I heard, Hugh was talking with the Keyspan people to get this resolved. But that was a few weeks ago.
Any news Hugh?
thanks,
greg k-h
Re: Fwd: Copyright infringement in linux/drivers/usb/serial/keys
Hi Greg,
Quoted 9 lines · On 25-May-2001 Greg KH wrote:
Quoted 7 lines · Aaron Lehmann
This message sparked a long thread on the debian-legal mailing list, which is long since dead. I am personally very curious about whether this has been resolved upstream. I consider it a very important issue, which is why I asked for RMS' opinion. He said that what is being done is clearly not "mere aggregation", and that such firmware should be moved out of the kernel (and even the tarball) to stop violating the GPL and make Linux be free software.
Quoted 4 lines
Last I heard, Hugh was talking with the Keyspan people to get this resolved. But that was a few weeks ago.
Any news Hugh?
Will follow it up.
Cheers,
Hugh
Not to sound dense, but what part of the GPL prohibits a piece of GPL'd software from including non-GPL'd code? The GPL does explicitly state that you can't include it's software in proprietary code, but I don't recall seeing a provision that prohibits the other way around.
It may not be in the "spirit" of the GPL, but as a legal document, the letter means more than the spirit in the final determination.
Sorry to intrude on this, but the thought just struck me. I could be wrong in my remembrance.
John
Quoted 4 lines
Not to sound dense, but what part of the GPL prohibits a piece of GPL'd software from including non-GPL'd code? The GPL does explicitly state that you can't include it's software in proprietary code, but I don't recall seeing a provision that prohibits the other way around.
The same thinbg holds true both ways. A work containing GPL code must be GPL or freer. Adding either to the other is the same thing.
The firmware is a seperate work, and its mere aggregation even if a .o file is a more unusual archive formt
Quoted 4 lines
which is why I asked for RMS' opinion. He said that what is being done is clearly not "mere aggregation", and that such firmware should be moved out of the kernel (and even the tarball) to stop violating the GPL and make Linux be free software.
Given that the firmware is a seperate work (try loading it under windows and you'll find its quite useful without Linux and does not depend on Linux) and that the folks I have raised this with - who should know their law - disagree with Adam I'm not currently persuaded.
The relationship between a USB driver and the kernel is no different to that between say a web browser and server chatting over a network. In this case the network is USB not IP based that is all. Clearly mozilla does not cease to be free software or a seperate work because you looked at microsoft.com with it.
Now if you want a plausible example of the kind of thing Adam is talking about you should take a hard look at the ACPI code, where ACPI bytecode is linked into the kernel by interpreting it
Alan