"Stolen" SCO Linux Code Snippets Leaked

stere0 writes "An article (in German) published on the German IT news site Heise includes two pictures (1, 2) of the "stolen" source code SCO claims to be theirs. Part of the first screenshot has been scrambled, the font has probably just been changed to Symbol; can anybody decipher it? I searched for the code snippets on Google. The code does indeed come from the kernel; the photographs show what seems to be lines 88-102 and 109-123 of /arch/ia64/sn/io/ate_utils.c from the 2.4 kernel tree. " Update: 08/19 16:39 GMT by M : LWN has a nice piece tracing the origins of the disputed code, and showing that SCO is simply lying.

Comments

Translation of "symbol" section:

[*igor*]

* As part of the kernel evolution toward modular naming, the
* functions malloc, and mfree are being renamed to rmalloc and rmfree.
* Compatibility will be maintained by the following assembly code:
* (also see mfree/rmfree below)
*/

Re:Translation of "symbol" section:

[Yohahn]

Others agree, they've been chatting about this on Linux weekly news:

see here

Re:Translation of "symbol" section:

[JanneM]

Um, no, that is a very premature conclusion.

First, they will have to show that this code is their property; as others have mentioned, both Linux and SCO Unix contain BSD code - which is perfectly legal to copy.

Second, if the code is stuff from IBM/Sequent, they will need to show that IBM did not in fact have the right to give the code away to Linux. This will not be determined until the lawsuit is settled (and looks unlikely to go in SCO:s favour considering the side agreements that explicitly give IBM permission to do whatever they want with their code).

And don't forget that in the case that IBM loses that suit, it becomes a matter between SCO and IBM only. That code IP is owned by IBM, not SCO, no matter what, and Linux users are free to continue using it. The only one damaged by that would be IBM, as they would be found for breach of contract.

Oh, and the 110/2000/890000/dozens/whatever number sco is flinging around at the moment is pretty much immaterial. They haven't exactly been paragons of accuracy in this affair so far.

Again, to reiterate, the whole "illegaly copied code" is, so far, just smoke and mirrors. There is no lawsuit alleging any copyright infringement. There is only a lawsuit alleging that IBM violated an agreement not to share some of their IP with third parties.

Re:Translation of "symbol" section:

[buffer-overflowed]

Since you probably didn't read some of the comments, as others have pointed out, these exact comments (and even some source) are in the public domained Unix 7 code, as well as old BSD code (BSD being absolutely in the clear from the settlement). So, this is in no way an example of infringment. Hell code I've wrote w/o ever seeing any of these is similarly constructed and commented.

BSD Comments (malloc), 1986:
/*
* Allocate 'size' units from the given map. Return the base of the
* allocated space. In a map, the addresses are increasing and the
* list is terminated by a 0 size.
*
* Algorithm is first-fit.
*/

SGI Comments, 1992 - Present:
/*
* Allocate 'size' units from the given map.
* Return the base of the allocated space.
* In a map, the addresses are increasing and the
* list is terminated by a 0 size.
* Algorithm is first-fit.
*/

Unix 7 (Public Domain) 1979:
/*
* Allocate 'size' units from the given
* map. Return the base of the allocated
* space.
* In a map, the addresses are increasing and the
* list is terminated by a 0 size.
* The core map unit is 64 bytes; the swap map unit
* is 512 bytes.
* Algorithm is first-fit.
*/

Wow, stunning proof, absolutely stunning that public domain source and comments would get used and modified.

Re:Translation of "symbol" section:

[Anonymous Coward]

> 7th Edition UNIX is NOT Public Domain.

Actually, it is:
http://linux.oreillynet.com/lpt/a/1595/

Re:Translation of "symbol" section:

[Anonymous Coward]

Actually, it's not.

The article says Caldera released V7 under the BSD license. That makes it free software but not public domain.

At any rate, it *does* mean that SCO's claims are groubdless.

Re:Translation of "symbol" section:

[Alsee]

Show me a case where you are allowed to keep stolen goods

Ah. You're apparently one of those people that thinks information is property and has absoutely no understanding of the LEGAL difference between copyright infringment and theft. Copyright infringement does not create any sort of "stolen goods" that need to be returned to the owner. The copyright holder is compensated with damages from the person who committed infringment, end of story.

Of course I expect you to go on a rant about how I'm "wrong", but I suggest you go argue with the US Supreme court:

"the rights of a copyright holder are `different' from the rights of owners of other kinds of property"

"the copyright holder owns only a bundle of intangible rights which can be infringed, but not stolen or converted"

"It follows that interference with copyright does not easily equate with theft, conversion or fraud. The Copyright Act even employs a separate term of art to define one who misappropriates a copyright: 'Anyone who violates any of the exclusive rights of the copyright owner,' that is, anyone who trespasses into his exclusive domain by using or authorizing the use of the copyrighted work in one of the five ways set forth in the statute, 'is an infringer of the copyright.'"

The US supreme court states that you CANNOT carry over "theft" concepts to a copyright infringement case. The US supreme court specificly specificly rejected the claim that infringing copies "involved stolen goods". See Dowling vs. United States.

-

Location in Sys 7

[Anonymous Coward]

See Sys 7 1979 location

Kernel mailing list comment

[Gaetano]

This from the kernel mailing list

http://www.tuhs.org/Archive/Caldera-license.pdf

January 23, 2002 Dear UNIX? enthusiasts, Caldera International, Inc. hereby grants a fee free license that includes the rights use, modify and distribute this named source code, including creating derived binary products created from the source code. The source code for which Caldera International, Inc. grants rights are limited to the following UNIX Operating Systems that operate on the 16-Bit PDP-11 CPU and early versions of the 32-Bit UNIX Operating System, with specific exclusion of UNIX System III and UNIX System V and successor operating systems: 32-bit 32V UNIX 16 bit UNIX Versions 1, 2, 3, 4, 5, 6, 7

-Tupshin

babelfished

[Empiric]

Here's a semi-readable, slightly-cleaned babelfish translation... it'd be great if somebody who can actually speak German could post a better one...


The fight for the legal standard of Linux develops more and more to the show piece: Still two weeks ago ago on the Linuxworld had itself boss Mathew Szulik as the rescuer of the free world explained and all Linux trailers to the fight against the chains of the commercial software industry called. Now geriert itself its opponent Darl McBride of SCO still more martialischer: a James bond in the struggle with dark power -- the open SOURCE movement.

SCO executive committee Darl McBride used two full hours for the prelude of the SCO forum, in order to represent the legal position of its company. With pictures and title music from James bond films the manager sought itself to join in the faithful ones of the former cult company from Santa Cruz for fight for property. The SCO Group leads a law case with IBM because of alleged copyright infringements and abuse of SCOs protected Unix program code in Linux. Star lawyer David Boies, which attained celebrity as a complaint representative of the US government against Microsoft, represents SCO IBM over 1500 Linux Grossanwender printing reminder approximately from SCO kept and was requested to pay royalties.

Supported of its vice-president Chris Sontag showed McBride of examples from the code of the Linux Kernelversionen 2,5 and 2,6, which are to prove that program sections were transferred invariably from Unix -- an example shown by SCO to code comments in the picture left ( version increased ). Identical typing errors in the comments as well as unusual ways of writing would have left traitorous traces, to stated Sontag. Around this to prove McBride a team for pattern recognition had angeheuert, around ten thousands from program lines to through forests. The few code sequences shown apart from the comments were made to a large extent illegible, alleged, in order to protect SCOs author-genuine. They would stand however representing for thousands of program lines, for stressed Sontag. From several persons or groups at different times parts were transferred illegaly to Linux and distributed sourceopen at users and developers. At the contentious software it goes besides not around simple or trivial functions, but important operating system characteristics for the fitness with fastidious tasks and in extremely safe operating conditions into enterprises. In addition belong the multi-processor mechanisms NUMA and SMP, which were to be had under Unix Lizenzbedingungen only with expensive hardware in the value of ten thousands from US dollar to.

Approximately 700 crucial code lines of the SMP technology are to have moved from Unix into the Linux releases 2,4 and 2,5. Altogether SCOs testers over 800.000 lines would have found duplicated program text -- an example of SCO shows the picture right ( version increased ). Attorney Mark Heise from the Boies boies-Kanzlei came along for the support of the SCO managers on the podium in read Vegas. It made clear that a GPL license did not protect against the requirement for authority of SCO. The Unix license, which bought SCO 1994 of the original Unix inventor RK & T, guarantees SCO property at Unix system v copyrights and all RKS & t-software and Sublizenzrechten. Originally the license agreement defined by RK & t-lawyers, which changed over by purchase to SCO, is clear in addition regarding the range and consequence of the license, stressed the lawyer. Afterwards the license grants the "right the software products to the licensee (for example IBM) to own business purposes to use internally", quoted Marks of Heise from the contract text. "modifications and derivatives of results are to be treated like the original software products", continue to be called it there. And they "cannot become used for others or by others".

"Now we know ourselves finally, like Linux in completely short time of a hobby operating system to the platform for ente

Re:IANAC

[Anonymous Coward]

The code in question in the Linux kernel is simply a function to allocate a block of space from a pool, looking for the first fit.

If an undergraduate experienced in C were asked to do the same problem the code would look very similar. Hardly a trade secret. Others have commented that this appears in earlier malloc libraries. Perhaps there's the common ancestry, way before SCO existed.

Hardly enterprise class stuff. They had better have much much better examples or their case is toast.

Re:I can decipher it!

[Fractal Law]

The thing is that it actually is the Windows Symbol font. I can read Greek (Ancient Greek at least) and while the alphabet used is the Greek one all that somebody did was highlight the text in question and change the font to Symbol, which is what Windows calls its Greek font.

In other words it's English written using the Greek alphabet. Why somebody would do something so silly puzzles me, however.

Yes, that's right, they're claiming malloc()

[raindog2]

Here's the earliest implementation people have found so far, from 1979 (before SCO was "born"):

http://minnie.tuhs.org/UnixTree/V7/usr/sys/sys/mal loc.c.html

And here's where it was part of BSD 2.11 circa 1992:

http://unix-archive.pdp11.org.ru/PDP-11/Trees/2.11 BSD/sys/sys/subr_rmap.c

Oh, how I hope the mainstream tech press "gets" this.

Code from BSD?

[Anonymous Coward]

I think that it looks like SCO's snippet, doesn't even belong to them, compare following which is: Copyright 1986 Regents of the University of California

That's BSD

Re:To sum up:

[Vexalith]

It's also in a 1986 Berkley file: http://unix-archive.pdp11.org.ru/PDP-11/Trees/2.11 BSD/sys/sys/subr_rmap.c

Re:I'm not the only one who noticed this...

[gotan]

There's also a very informative lkml thread about this and it's already been removed from the source tree, but apparently not because of copyright issues but because it was just "ugly as hell".

Re:Code in picture 2 doesn't even compile

[PiGuy]

Here's the actual kernel code:

if (size == 0)
return((ulong_t) NULL);

Now, where'd that misteak come from?

This code was apparently donated by Caldara (SCO)

[m.dillon]

From the linux-kernel list, the code was apparently donated by Caldara under the BSD license in 2002. Here are the references.

Start of Thread

Conclusion

It's HP's fault this stupid code is in there.

[echo]

Check this out

patch@hp.com according to bitkeeper.

Also, this has been removed in 2.6, mainly because it was a stupid implementation.

Re:To sum up:

[Vexalith]

Replying to myself with a link:
http://oasis.dit.upm.es/~jantonio/documento s/opens ource/kirkmck.html

LWN Overview

[saspengiun]

Check out the history good overview History

Code has been around since at least 1973

[rkww]

The nsys kernel version of malloc has exactly the same implementation, albeit without any comments.

Dennis Ritchie has written So far as I can determine, this is the earliest version of Unix that currently exists in machine-readable form. ... The dates on the transcription are hard to interpret correctly; if my program that interprets the image are correct, the files were last touched on 22 Jan, 1973. ...

Re:oh no!

[Arker]

Anyone ever given serious thought that perhaps the SCO code was lifted from the Linux source? The SCO engineers obviously had the same level of access as everyone else in the public domain -- perhpas the case will come down to who can prove what was released when. .anacron

That may be the case in many of their claimed 'infringing lines' but not in these two.

The first one shows comments (not code) which match. They're also straightforward descriptions of functionality, date back at least to 1979, and occur in a number of old unix versions which have been commonly read and used for teaching purposes for years, as well as the Lions book.

In this case the comment definately predates Linux so couldn't have been copied from Linux, but the fact it occurs in Linux code is not strong evidence of copyright infringement.

http://minnie.tuhs.org/UnixTree/32VKern/usr/src/ sys/sys/malloc.c.html
http://minnie.tuhs.org/UnixTree/Interdata_v6/usr/s ys/malloc.c.html
http://minnie.tuhs.org/UnixTree/Ausam/sys/ken/mall oc.c.html

The second section of code is copyright of and contributed by SGI, and concerns functionality that SysV doesn't have. It's only relevant on the assumption that SCOs theory that it owns every technology any Unix licensee ever added to their own version of Unix holds water, which means only if the court winds up throwing the entirety of copyright law and precedent out the window and writing new law just for SCOs benefit.

Translation of the article

[reignbow]

The battle for the legality of Linux is becoming increasingly melodramatic: Two weeks ago, RedHat CEO Mathew Szulik declared himself Saviour of the Free World, and called for all Linux-Supporters to join the battle for freedom from the software industry. His opponent darl McBride from SCO strikes an even more martialic pose: a James Bond fighting against the forces of Darkness -- incarnated in the Open Source Movement.

McBride spent no less than two hours at the beginnign of the SCO symposium to clarify his company's legal position. Backed by pictures and music from various Bond flicks, he attempted to rally the supporters of the formerly cult company from Santa Cruz to his fight for the Good Cause. The SCO Group has started a legal battle against IBM for alleged copyright violations and misuse of SCO-owned UNIX code in Linux. Star attorney David Boies, famous for fielding the US anti-trust effort against Microsoft, represents SCO against IBM. More than 1500 major Linux-using companies have received admonitions to pay licensing fees from SCO.

With Vice CEO Chris Sontag as sidekick, McBride offered several examples from Kernel 2.5 and 2.6 that are meant to prove that several program parts were transplanted unmodified from UNIX -- such an example is here. Duplicated typos in the commentary as well as unusual coding style have left traces, says Sontag. To porve this, McBride employed teams for pattern recognition to parse tens of thousands of lines of code. The few sequences of actual code shown besides the commentaries were largely scrambled, supposedly to protect SCO copyright. They were, however, representative for a thousand other just like them, emphasizes Sontag. Multiple developers had illegally transplanted code into Linux and then distributed the source to users and developers. The software in question is nothing trivial, but contains integral operating system functions used for demanding applications and extremely secure environments in companies. Among them are the multiprocessor technologies NUMA and SMP, which under UNIX licensing cost 10,000$ or more.

Rouhgly 700 lines of code for the SMP technology are supposed to have gone into Kernel versions 2.4 and 2.5. All in all, SCO claims to have found no less than 800.000 lines of duplicated code -- one example is shown here. Attorney Mark Heise from Boies' law firm joined the SCO chiefs on the podium in Las Vegas. He emphasized that the GPL did not offer protection against copyright claims from SCO. The Unix license that SCO bought from AT&T in 1994 guarantees SCO ownership of System V copyright and all AT&T software and sublicense rights. The license agreement, originally drawn up by AT&T lawyers, which has since gone over to SCO, is unequivocal concerning scope, Heise affirmed. Accordingly, the license gives the licensee (e.g. IBM) the right to use the software internally for commercial purposes. Modifications and derivatives are subject to the license just like the original. They cannot be used for or by third parties.

"Now we finally know how Linux has matured from hobby OS to IT-company platform," Sontag jibes. "If something sounds too good to be true, it usually isn't," topped McBride. Evolved technology simply cannot be had for free. "Free Software -- not our thing." UNXIX comprises 20 years of development work: Based on it, SCO wants to make money for another 20 years. McBride appealed for support from partners and developers from the UNIX community, otherwise, "the times for good business might soon be over." GPL and Open Source destroy legal business models -- compensations and a legal business model for the future are therefore necessary. Heise seconded: That SCO once distributed its code as Linux distributor, did not mean that Linux users where protected from all demands because of the GPL. Copyright for code can only be obtained by a written contract wit

Well

[autopr0n]

If you read the article, you'd see that that SCO BSD'd all that code just last year. This is something they can't revoke.

Oops missed the best one:

[Arker]

http://unix-archive.pdp11.org.ru/PDP-11/Trees/2. 11BSD/sys/sys/subr_rmap.c

Deliberately not making links so as to hopefully not slashdot more servers than necessary. If you want to see it, cut and paste.

BSD 2.11, for the PDP-11, had it. This is very very ancient Unix.

Re:oh no!

[Royster]

The last link shows the code apparently came from SGI. Isn't SCO claiming that IBM contributed the so-called infringing code?

No.

The accusations made by SCO have been very unclear. The IBM case is about acts by IBM which SCO claims breaches the IBM/AT&T contract. It involves contributing code IBM got by buying Sequent and by participating in Project Monterey. This technology includes RCU, NUMA, SMP and scalability.

Independantly, SCO claims that there are many lines of code copied from Sys V into Linux. They have not yet filed any copyright suit against anyone since they only just got the copyright registrations issued.

Don't confuse the two very different legal issues.

Analysis by Bruce Perens

[kaip]

Bruce Perens has written an analysis of the code that SCO claims was wrongfully copied into Linux: http://perens.com/Articles/SCOCopiedCode.html

Re:Legal summary of the SCO situation

[frkiii]

Copyright infringement is NOT part of their suit against IBM. The only place where SCO has alledged this, is in their statements to the press, in the form of "IP", etc. To the best of my knowledge, they have not filed a copyright infringement case against ANYONE. I have read SCO's original and amended suits against IBM, it is simply a contractual suit, generally breach of contract, and disclosing "trade secrets". Which is really funny, because they specifically name NUMA, JFS and SMP in their lawsuit, refering to what IBM contributed to Linux. And that is funny, because SCO does not own or "control" that software at all. Most if it is copyrighted by IBM or by company that IBM bought (Sequent) or others, but NOT SCO. SCO's entire case, based on their filed amended complaint, hinges on the court ruling that NUMA, JFS and SMP are "derivative workds" of Unix System V. But that is very highly unlikely, in my opinion. Regards, Fredrick