| 1 |
On Tue, Dec 29, 2009 at 09:42:06PM -0500, Vincent Launchbury wrote: |
| 2 |
> Greg KH wrote: |
| 3 |
> > The fact that some people claim that the firmware blobs somehow violate |
| 4 |
> > the GPLv2 license of the kernel is a claim, not a fact, so please do not |
| 5 |
> > state it as such. |
| 6 |
> |
| 7 |
> Hi Greg, |
| 8 |
> |
| 9 |
> Thanks for your reply. |
| 10 |
> |
| 11 |
> I think you misunderstood my point though. I wasn't saying that the |
| 12 |
> firmware violates the GPL, I have no idea whether it does or not. I was |
| 13 |
> saying that some of the firmware is non-free software, and therefore the |
| 14 |
> license should include more than just GPL-2. This especially effects |
| 15 |
> people using ACCEPT_LICENSE to maintain a free system. |
| 16 |
|
| 17 |
Heh, no, it does not, unless your BIOS, and your keyboard firmware, and |
| 18 |
your mouse firmware are all under a "free" license. The only thing |
| 19 |
close to this type of machine is the OLPC, and even then, I don't think |
| 20 |
all the microcode for the box was ever released. |
| 21 |
|
| 22 |
So it's a pointless effort. |
| 23 |
|
| 24 |
Hint, these firmware blobs do not run on your processor, so they have |
| 25 |
nothing to do with the license of your "system". |
| 26 |
|
| 27 |
> > Also note that the majority of these firmware blobs are now removed |
| 28 |
> > from the kernel, and are in a separate patckage, so this might be |
| 29 |
> > totally irrelevant at this point in time. |
| 30 |
> |
| 31 |
> This may be true, but the packages in the main tree still contain |
| 32 |
> non-free firmware. If this is fixed in a later release, then GPL-2 would |
| 33 |
> be fine for those. |
| 34 |
|
| 35 |
Again, no, the GPLv2 covers the license of all of the code you run in |
| 36 |
the kernel package. |
| 37 |
|
| 38 |
> > So please don't state that the Linux kernel is not properly listed as |
| 39 |
> > the GPLv2, because it is. |
| 40 |
> |
| 41 |
> In linux-2.6.31 for example, here are some excerpts from |
| 42 |
> firmware/WHENCE: |
| 43 |
> |
| 44 |
> Regarding the keyspan USB driver: |
| 45 |
> This firmware may not be modified and may only be used with |
| 46 |
> Keyspan hardware. |
| 47 |
> |
| 48 |
> and the emi26 driver: |
| 49 |
> This firmware may not be modified and may only be used with the |
| 50 |
> Emagic EMI 2|6 Audio Interface. |
| 51 |
|
| 52 |
And again, you do not run those firmware images on your processor, so |
| 53 |
the point is moot. |
| 54 |
|
| 55 |
And note, _I_ placed those images in the kernel image, after consulting |
| 56 |
lawyers about this issue, so it's not like I don't know what I am |
| 57 |
talking about here. |
| 58 |
|
| 59 |
> I'm not sure if this git repo is part of a separate package or not, but |
| 60 |
> it seems the same terms are present: |
| 61 |
> http://git.kernel.org/?p=linux/kernel/git/dwmw2/linux-firmware.git;a=blob;f=WHENCE;h=83d245bee1ec44cbd5c0e1a53a3989c57f675c91;hb=f20b0674534a444ae74239843cac19f72c64912b |
| 62 |
> |
| 63 |
> Which is why I think the license should be amended. If I'm mistaken, |
| 64 |
> please do correct me, but based on my above notes, I believe it should |
| 65 |
> be updated. |
| 66 |
|
| 67 |
Please see above why this is all just fine. |
| 68 |
|
| 69 |
thanks, |
| 70 |
|
| 71 |
greg k-h |
Archives