AMD Releases "Open-Source" UVD Video Support
- Login o registrati per inviare commenti
I am a member!
Does anybody know if this means a 100% free 3d driver will work? Apparently this video acceleration has been a problem for the same reasons the driver isn't in Trisquel (digital restrictions). What I'm curious is if the "open-source" driver will now work without any non-free bits.
http://www.phoronix.com/scan.php?page=article&item=amd_opensource_uvd&num=1
I am a member!
It appears not. The blurb "Update: The code just arrived!" links to a mailing list post, which then links to "the newest firmware" which includes this restriction: "No reverse engineering, decompilation, or disassembly of this Software is permitted."
I am a member!
I don't think so either. However I don't know enough about whats holding things back. If those non-free bits that exist are not critical or the legal review has free'd up enough of the restrictions maybe it can be done. I'm pretty sure the DRM was the reason given that we don't have a free driver. I forget if this is the old reason or if I had updated info on the real reason being elsewhere. I did get third party info from someone with a different reason than the official one.
That's too bad. If there were a company that were freedom friendly(like ThinkPenguin) producing a graphics card that had reasonable specs and worked without non-free firmware and drivers I think that a lot of people would buy it, even if it were a somewhat inferior card. I know that someone was trying to produce an "open" graphics card a few years back but I don't think it ever went to far.
On 03/04/13 14:02, Chris wrote:
> Does anybody know if this means a 100% free 3d driver will work?
I looked up UVD and found this:
https://en.wikipedia.org/wiki/Unified_Video_Decoder
It doesn't sound like it, but I don't know too much about GPUs.
Someone posted the following in the comments of Slashdot, claiming it is the license of this 'code:'
---------------
REDISTRIBUTION: Permission is hereby granted, free of any license fees, to any person obtaining a copy of this microcode (the "Software"), to
install, reproduce, copy and distribute copies,
---> in binary form only, <---
of the Software and to permit persons to whom the Software is provided to do the same, provided that the following conditions are met:
No reverse engineering, decompilation, or disassembly of this Software is permitted.
Redistributions must reproduce the above copyright notice, this permission notice, and the following disclaimers and notices in the Software documentation and/or other materials provided with the Software.
---------------
It's not source code. It's firmware. This link:
http://lists.freedesktop.org/archives/dri-devel/2013-April/036766.html
contains a link to:
http://people.freedesktop.org/~agd5f/radeon_ucode/
which is ALL firmware. I recognize it because it could be built into the kernel on Gentoo.
I am a member!
What does not make sense to me is why did they need to do a legal review to release non-free binary bits? They wouldn't be revealing anything secretive in non-free code. In fact doesn't the non-free driver already support it? This article doesn't really make much sense to me.
They probably needed the review to release the kernel and mesa driver
parts. There is new UVD firmware (215 KiB for one chip, previously the
biggest file was 32 KiB) and updated RLC firmware, maybe the
availability of much more code will make reverse engineering easier.
DRM isn't their only argument for not releasing sources nor
documentation of this firmware, the resources needing for the legal
review of the documentation are another reason that they state.
I am a translator!
It sees like AMD can't support or release anything "open source" without half-assing the thing.
I hear AMD have been helpful in regards to BIOS though. [1] They've also released useful GPU specs and taken part in the driver development, although their firmware remains proprietary.
Intel has the only manufacturer supported GPU (firmware+driver) but their wifi won't work.
nVidia couldn't care less about freedom. (But no thanks to them, a reverse engineered firmware+driver exists)
Intel is also the only manufacturer making their GPU drivers work
without a nonfree VBIOS (after the system is booted, although there is
work on a replacement for Coreboot). AMD machines with a mostly free
BIOS need a nonfree VBIOS for the KMS and xf86-video-ati drivers to work
without acceleration (this is the only blob in the YeeLoong 8133 boot
firmware).
Google working in coreboot for Chromebook:
Google Continues Working A Lot On Coreboot:
http://www.phoronix.com/scan.php?page=news_item&px=MTI3ODA
Google Provides Coreboot For Chromebook Pixel:
http://www.phoronix.com/scan.php?page=news_item&px=MTMxMDI
Google Does Sandy/Ivy Bridge In Coreboot For Chrome OS:
http://www.phoronix.com/scan.php?page=news_item&px=MTA4Mjg
Google demos Coreboot on Chromebook prototype, hints at Ivy Bridge support (video): http://www.engadget.com/2012/04/13/google-demos-coreboot-on-chromebook-prototype-hints-at-ivy-brid/
aka Samsung Lumpy aka chromebook:
http://www.coreboot.org/Samsung_XE550C22-H02US
coreboot ARM SoC's mainboard port:
http://www.coreboot.org/Project_Ideas#coreboot_ARM_SoC.27s_mainboard_port
On one hand, its good that a company (Google) that is making decent hardware (Chromebook Pixel) is supporting Coreboot. On the other, its a shame that the major hardware providers are still Microsoft's slave and would piss off Microsoft if they implemented it.
I am a translator!
nVidia did release a free software (but obfuscated) 2d only driver called "nv" a few years ago.
I guess I must fortunately admit I'm partially wrong about nvidia.
http://www.phoronix.com/scan.php?page=article&item=nvidia_tegra_3d
- Login o registrati per inviare commenti