GNU Linux-libre 6.11 Makes Adaptations For Rust, Warns Of Hidden Binary Bits In v6.11
([GNU] 2 Hours Ago
GNU Linux-libre 6.11-gnu)
- Reference: 0001491958
- News link: https://www.phoronix.com/news/GNU-Linux-Libre-6.11
- Source link:
Building off yesterday's [1]Linux 6.11 release , the GNU Linux-libre 6.11-gnu kernel is now available that is the downstream stripping out driver support/features depending upon closed-source microcode/firmware and other modifications in the name of software freedom and ensuring no closed-source bits are used on Linux-libre-enabled systems.
GNU Linux-libre 6.11 has continued cleaning and de-blobbing various drivers such as AMDGPU graphics, various network drivers, the Intel AtomISP driver, and others. Plus there's now detection for firmware-loading primitives within Rust programming language code. The GNU Linux-libre 6.11 highlights include:
- Cleaned up amdgpu isp, tn40, rtl8192du, cs40l50, rt1320, pcie-rcar-gen4, and new AArch64 devicetree files.
- Updated cleaning up of amdgpu, adreno, vgxy61, atomisp, btnxpuart, prueth, tas2781.
- Added nowait-nowarn version of the firmware loading/rejecting primitives.
- Added detection of firmware-loading primitives in Rust.
And in usual fashion they have released a new graphic of their "Freedo" mascot for Linux-libre 6.11:
They also warned of new code in Linux 6.11 that is a binary sequence embedded within new GPLv2 driver code:
# Bad news
When preparing 6.11-rc7-gnu, tutor Alex caught a long sequence of numbers that looked suspicious, contacted the contributor of the new driver, and got confirmation that it was indeed executable code generated from unreleased source code.
Back when upstream started adding binary blobs to Linux, they were argued to be separate programs, under different licenses. But this new one was contributed explicitly under the GNU GPL, so, as long as its source code remained unavailable, it could not be distributed at all.
The contributor agreed, and immediately proceeded to address the legal problem, posting patches to move the blob out. Unfortunately, the top maintainer of Linux doesn't seem to have seen a problem there, and proceeded to the final release without the fix.
IANAL, but even as a penguin I understand that distributing it that way fails to comply with the requirements of the stated license, and can thus get your license automatically terminated. So, those of you who plan on distributing that release, watch out and stay out of trouble!
Even more unfortunately, a deeper problem remains: the program that must be loaded onto the device for it to work remains nonfree, so the device remains incompatible with the freedoms that all users deserve.
# Good news
On the good side, that binary blob can still be obtained from upstream under the GNU GPLv2, so anyone interested in it can proceed to legally reverse engineer it, and get us all one more Free piece of firmware.
More details within the [2]release announcement and [3]FSFLA.org for its project site.
[1] https://www.phoronix.com/news/Linux-6.11
[2] https://lists.gnu.org/archive/html/info-gnu/2024-09/msg00007.html
[3] https://www.fsfla.org/ikiwiki/selibre/linux-libre/#news
GNU Linux-libre 6.11 has continued cleaning and de-blobbing various drivers such as AMDGPU graphics, various network drivers, the Intel AtomISP driver, and others. Plus there's now detection for firmware-loading primitives within Rust programming language code. The GNU Linux-libre 6.11 highlights include:
- Cleaned up amdgpu isp, tn40, rtl8192du, cs40l50, rt1320, pcie-rcar-gen4, and new AArch64 devicetree files.
- Updated cleaning up of amdgpu, adreno, vgxy61, atomisp, btnxpuart, prueth, tas2781.
- Added nowait-nowarn version of the firmware loading/rejecting primitives.
- Added detection of firmware-loading primitives in Rust.
And in usual fashion they have released a new graphic of their "Freedo" mascot for Linux-libre 6.11:
They also warned of new code in Linux 6.11 that is a binary sequence embedded within new GPLv2 driver code:
# Bad news
When preparing 6.11-rc7-gnu, tutor Alex caught a long sequence of numbers that looked suspicious, contacted the contributor of the new driver, and got confirmation that it was indeed executable code generated from unreleased source code.
Back when upstream started adding binary blobs to Linux, they were argued to be separate programs, under different licenses. But this new one was contributed explicitly under the GNU GPL, so, as long as its source code remained unavailable, it could not be distributed at all.
The contributor agreed, and immediately proceeded to address the legal problem, posting patches to move the blob out. Unfortunately, the top maintainer of Linux doesn't seem to have seen a problem there, and proceeded to the final release without the fix.
IANAL, but even as a penguin I understand that distributing it that way fails to comply with the requirements of the stated license, and can thus get your license automatically terminated. So, those of you who plan on distributing that release, watch out and stay out of trouble!
Even more unfortunately, a deeper problem remains: the program that must be loaded onto the device for it to work remains nonfree, so the device remains incompatible with the freedoms that all users deserve.
# Good news
On the good side, that binary blob can still be obtained from upstream under the GNU GPLv2, so anyone interested in it can proceed to legally reverse engineer it, and get us all one more Free piece of firmware.
More details within the [2]release announcement and [3]FSFLA.org for its project site.
[1] https://www.phoronix.com/news/Linux-6.11
[2] https://lists.gnu.org/archive/html/info-gnu/2024-09/msg00007.html
[3] https://www.fsfla.org/ikiwiki/selibre/linux-libre/#news
kurkosdr