GPL Issues Surrounding Commercial Device Drivers?
Demiurg asks: "My company has recently decided to support Linux for it's embedded networking products which means that I'm starting to write Linux device drivers for our hardware. The company was very concerned about GPL issues and consulted a lawyer - who advised us to go for a user-space driver, saying that this is the only safe way to avoid GPL issues. I tried to give them a few examples of companies distributing binary only drivers (NVIDIA and Rational) but was told that these companies do not distribute binary only drivers - they only allow you to download them from a web site (which is not an option for an embedded product). What does Slashdot have to say about the issue? Is writing a user-mode (and hence not very efficient) driver the only way for a company to protect it's intellectual property? Please refrain from giving answers like 'all code should be GPL' - although I personally may agree, such answers will not help me convince management to make the change." Are there any lawyers (or readers with the right legal knowledge) out there that can confirm or contradict this recommendation?
So, you got your advice from a legal consul, and you're thinking about ignoring it in favor of advice from the /. crowd? Is that smart?
Je ne parle pas francais.
It all boils down to the concept of "derivative work". Is a device driver a derivative work of the kernel? My opinion is no, but only the courts can truly answer that question, and no one has asked them yet.
And the men who hold high places must be the ones who start
To mold a new reality... closer to the heart
Why is a userland driver on linux necessarily slower? Shouldn't a well-written user space driver at an appropriate priority level be just as fast as a kernel driver? And if not, why is this slow on linux but fast on some other operating systems, and what can be done to fix it?
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
Seems like you might get good info by directly talking to the FSF or linux kernel mailing list... I'm sure some people would piss and moan about binary this and proprietary that, but in the end I think you'd find more people eager to help expand linux's reach in an optimally efficient manner. IIRC the FSF's head legal eagle is a guy named Eben Moglen (there was something posted on /. just yesterday that involved his name, again IIRC.) Personally, although IANAL, making something available for public download from a website sounds pretty much like "distributing" something. True, the user has to decide to download it, but then again your users would have to decide to buy your particular widget (thus potentially getting the linux driver) also...
News for Geeks in Austin, TX
Nvidia and others do "distribute" binary-only drivers. Just because they don't actively send out CD's with their drivers does not mean they do not distribute them. Under the GPL both methods are concidered a distribution. As far as writing drivers, as long you use the driver module method and do not need to directly change any preexisting kernel code, I do not believe there is any problem at all. That is how Nvidia and others can get away with that.
You need to look at exactly what parts of your hardware a GPL'd driver source would reveal, and whether there is really any need to avoid a GPL'd driver for this reason.
Only the people who know what the hardware does and why the features in question need to be kept secret can make this decision.
-- BtB
John Sauter (J_Sauter@Empire.Net)
So you your company should go for the best possible driver, to increase hardware sales. If that means GPL'ing the driver, you can also benefit from the expertise of a lot of developers.
Ask your lawyer how distributing something over the internet is not distributing?
NVidia are distributing binary-only drivers. They do a trick however, where the code that needs to be loaded into the kernel is (L)GPL, and then this cpde interfaces their binary driver. This both allows a buffer to the GPL code but also to support various different versions of the kernel with the same driver.
Moreover it allows for a lot of naive trolls claiming that Nvidia is distribting the source-code because they have seen this little module.
Why is a userland driver on linux necessarily slower?
See subject.
And if not, why is this slow on linux but fast on some other operating systems, and what can be done to fix it?
No operating system can have userspace drivers without context switches. Some OSes have very low context switch times and *do* put all their drivers in userspace (which has a number of very big advantages I'm not going into right now) -- but they still, unavoidably, pay some performance penalty for doing so.
The fact that NVIDIA and Rational have not been sued does not imply that they are legally in the clear. There are certainly several kernel developers who have said clearly that the "module exception" invented by Linus Torvalds does not extend to linking to their code. So far they have not sued anyone. Maybe they never will.
Finally! A year of moderation! Ready for 2019?
In general, do not take advice on how to write a device driver from a laywer. Or, if he/she insists, invite them to write the driver for you. Pay only if it works.
Have you got your LWN subscription yet?
I don't quite see what is lost by releasing a device driver under GPL.
:)
You get:
1) an army of geeks who will instantly worship your company and buy your products;
2) a slightly smaller army of geeks who can actually help you fix bugs in your own product FOR FREE.
You lose... erm, nothing?
The only drawback is that the interface to your hardware (register maps etc.) are no longer as secret as they were. Perhaps you're concerned about competitors stealing your ideas based on that knowledge? Well, think about it this way - if they badly want that information, they will already be reverse engineering your products and your binary-only drivers with a debugger and some elbow-grease.
IMO - you have nothing to lose. Embrace the future...
These sigs are more interesting tha
No operating system can have userspace drivers without context switches.
By that I mean no operating system with memory protection for userspace apps. It's possible to give up this benefit (even selectively) for improved userspace performance; see Kernel Mode Linux for a kernel patch that does just that (and which might be applicable to this fellow's situation).
Easy. People ask slashdot because slashdot readers have actually had to deal with these matters on a day to day basis. Your average lawyer may see a handful of case studies.
Now, he may have gone to the sort of lawyer that specializes in software licenses (what sort of lawyer is this? a very very rich sort), in which case you're right that it is kind of stupid to approach the trolls and IANALs. But if he went to Joe Q. Publicstein of Publicstein, Stern and Lowe, chances are he'll get a much more specific answer and leads on a case he may be able to point his legal eagle towards. Which could result in a much more informed decision.
By the by, I've got a feeling that the people who MIGHT press a case on the GPL have bigger fish to fry than a company writing a device driver. If they can't catch Microsoft biting GPL code, why are they going to hunt down developers expanding the OSS market?
Hey freaks: now you're ju
The kernel exports certain symbols that are available for driver modules to link to. These symbols are classified by license. Many symbols a driver can link to without having to be a GPL'ed driver. These symbols, presumably, are written by people who grant permission to link against non-GPLed code. Other symbols (for example the new work-queues stuff) are exported as so-called GPL_ONLY symbols, and linking against those in a non-GPL module is illegal. Thus, if you avoid using the GPL_ONLY symbols you can probably get away with a binary only module. The fact that this mechanism exists implies permission, so this aspect would be important to bring up to the lawyer.
A deep unwavering belief is a sure sign you're missing something...
If identifiable sections of that work are not derived from the Program, and can be reasonably considered independent and separate works in themselves, then this License, and its terms, do not apply to those sections when you distribute them as separate works. But when you distribute the same sections as part of a whole which is a work based on the Program, the distribution of the whole must be on the terms of this License, whose permissions for other licensees extend to the entire whole, and thus to each and every part regardless of who wrote it.
I think the keys here are what count as distribution and separate works. Obviously, the GPL was drafted with traditional software in mind and is not tailored toward embedded systems. You have to answer the questions "Is this a work based on the Program [a piece of software covered by the GPL]?" and "Is this an independent and separate work?" Unfortunately, I don't have enough facts about your system, and we don't have an attorney-client relationship, so I can't give you an answer. What I can tell you is that you should make sure your lawyer is familiar with your code development processes and distributions systems and knows enough about software development so that he can draw principled distinctions between different ways of doing things.
Laws affecting technology will always be bad until enough techies become lawyers.
My gut feeling is that a loadable module that does not itself incorporate any GPL'd code does not fall under the GPL just because it's run in conjunction with a GPL'd kernel. After all, even if the kernel + driver are considered together to create a derivative work, it's the user, and not you, who creates that derivative work at runtime.
And, the fact that you're using GCC to compile the module doesn't mean anything in and of itself. The real question is the license status of any libs that are linked into the resulting object code file. Whether they are under GPL, LGPL, or something else makes a big difference.
Now for the shades of gray part... any time you're dealing with GPL questions, you have to look at the genealogy of the running executable, working backwards to see what licenses govern the component pieces. It's a very fact-intensive process, and without understanding where all the code came from, you can't come close to giving a reasonable answer.