Linux 5.9 Brings Safeguard Following NVIDIA's Recent "GPL Condom" Incident

[L_A_G]

I'm not sure being very clearly antagonistic towards Nvidia by referring to this as "taint" in the code is the right way to go about this...

Sure, I understand that people who hate closed source software with a passion are going to hate shims like this to avoid having to open source stuff, but the choice here isn't between open and closed source. It's a choice between functionality and no functionality. Being proud of this is like being proud over not using energy wasting incandescent light bulbs whilst sitting in the dark because you don't have any LED or other more energy-efficient light bulbs to use instead.

I'm not saying that we shouldn't try to push Nvidia towards open-sourcing more of their stack, but we need to avoid needlessly antagonizing them and to allow them to do it in baby steps, encouraging them to do more each step of the way. There may be more catharsis in just out and out bashing them like Linus at Aalto University (then TKK) years ago or like now by referring to their code as "taint" because it's trying to avoid having to put pretty much their whole stack under full GPLv2. However all that achieves is discourage them from open sourcing more of their currently closed source software.

[JMB9]

Of course it's all OMG AMD OPEN SOURCE RULEZ, NVIDIA CLOSED SOURCE SUCKZ until you start working with a product and realize that Open Source means crap in terms of proper support.

This is nonsense. Yes, you do have valid arguments - but this conclusion can not be further from the truth.
AMD is still getting ground - and I expect AMD to invest much clearer in people writing code for Linux and Mesa in future.
Currently they opened up - but write code for a proprietary mix of Windows and Linux and trying to open up the Linux code as they can.
It is still more similar to the way Nvidia works - but without Nvidia trying to open their code at all.

Concerning professional Linux support it is not doable without having the source code (the reason for the tainted flag - no proper support possible) - so this statement is total madness.
Otherwise you have to play ping pong with all participants bringing in proprietary drivers and this is not fun (been there) - and an extreme waste of time.
So even though felt quality of users of proprietary code may be OK - I know several examples of games not running on Nvidia but even on iGPUs (and of cause on AMD dGPUs).

And concerning stability - I am working on AMD Navi 10 - with Kernel 5.7.15 and Mesa 20.1.5 - without any problem.
There are still missing parts (ROCm or a real configuration tool) ... but it is on it's way. And it is the 1st incarnation of RDNA - so coming Navi2 will be much easier to get up and running (and a better choice for problemless entry).
And here the open source drivers are superior to the shared (Linux/Windows) drivers - which should be expected - and seen.

Though totally off topic - sorry for that - but such nonsense is not to be left without a salt of truth!

And using Polaris right now is - well - not that technically reasonable (FullHD should be over more than 5 years ago for professionals).
It is an old design ... and even Vega is very limited.
So I waited quite long to be able to purchase non-Intel HW to be able to fill my future needs.
And I am really curious if AMD got things right - increasing efficiency as announced - and be more than competitive with Navi2.

[torsionbar28]

Y'all are missing the key phrase of birdie's post -- RX 5600 XT. It's a known thing among the AMD Linux crowd that the newest AMD GPUs usually have bad to horrible Linux support for up to a year, sometimes more. A Linux user, possibly into gaming, swapping out a new AMD card for a new Nvidia card shouldn't surprise a whole lot of people.

Also the part where there is a proprietary AMDGPU-PRO driver that is not FOSS but tends to be a bit ahead in terms of compatibility and stability. Not sure why birdie would bother with changing hardware, seems like poor troubleshooting on his part. My Vega56 card is rock solid stable on Ubuntu 20:04 for Steam gaming, and it does all the things you would expect, like video decoding and encoding, etc.

[JMB9]

Sanity is not strong with this one. You may want to look up in the dictionary what these words actually mean. Open Source fans are exactly the reason why many people skedaddle from everything related to Open Source.

Well, I am sorry - I thought about technically skilled people here.
Malicious means bad in a sense of backdoors, doing things the owner of the devices does not want etc.
And contamination means spoiling e.g. fresh water with oil (or poison).
So in this picture you have the audited and inspected Linux code and than putting in binary code which may do things which are not wanted similar to poison in water.
I am still sorry for you and I hope it helps.
And no - Starwars should only be cited by people interested in sciences.

[shmerl]

I prefer to use a GPU which has high quality drivers

And today that's AMD and Intel. Nvidia's drivers are anything but high quality. It took them decades to add PRIME support and even today it doesn't work properly. Shilling for Nvidia won't change the facts.

[birdie]

Well, I am sorry - I thought about technically skilled people here.
Malicious means bad in a sense of backdoors, doing things the owner of the devices does not want etc.
And contamination means spoiling e.g. fresh water with oil (or poison).
So in this picture you have the audited and inspected Linux code and than putting in binary code which may do things which are not wanted similar to poison in water.
I am still sorry for you and I hope it helps.
And no - Starwars should only be cited by people interested in sciences.

Not a single person on Earth is able of auditing the entire source code of the Linux kernel and proving it does not have backdoors or malicious functions. Again, sanity is not strong with this one. You're absolutely delusional in pretty much everything you're saying. Also, you don't follow prominent Linux security researchers (e.g. https://twitter.com/dvyukov ) and companies auditing the Linux source code who discover various serious errors daily despite the code being "open". Also, you're not aware that Linux kernel developers have recently stopped accepting various coding errors in order to fix them and their current stance is, "If you've found an error, send us a patch" (can't readily find the message - it was a few months ago).

You're so full of BS it's actually cringeworthy. Also, there hasn't been a single instance of NVIDIA being implicated in planting backdoors or intended "bad" code for the entire history of the company. You know unlike you, who's mostly spreading unsubstantiated BS under a generic nickname, the company has a reputation to keep.

TLDR: you have no proofs at all that NVIDIA has any malicious code in their closed source drivers, and you have no proofs open source code is free from malicious code. In short you're a BS'er.

And today that's AMD and Intel. Nvidia's drivers are anything but high quality. It took them decades to add PRIME support and even today it doesn't work properly. Shilling for Nvidia won't change the facts.

Buying an AMD CPU along with an AMD GPU and finding critical bugs (shown in the thread earlier) in the AMD open source drivers is shilling for NVIDIA. Right. Open Source fans have no common sense and logic.

[ermo]

I'm not sure being very clearly antagonistic towards Nvidia by referring to this as "taint" in the code is the right way to go about this...

From a kernel development perspective, being able to set a flag that clearly says "this kernel is considered tainted because it runs with non-GPL modules to which we do not have source code access" is a benefit in that the Vendor is clearly responsible for making their code work with Linux, not the other way around -- particularly when troubleshooting an issue.

This has been the philosophy from day one as I understand it.

Besides, Linux and NVidia have both been around for so long that NVidia really has no excuse for pretending that it doesn't understand how Linux kernel development works. The same goes for Facebook and the developer in question.

[birdie]

Also the part where there is a proprietary AMDGPU-PRO driver that is not FOSS but tends to be a bit ahead in terms of compatibility and stability. Not sure why birdie would bother with changing hardware, seems like poor troubleshooting on his part. My Vega56 card is rock solid stable on Ubuntu 20:04 for Steam gaming, and it does all the things you would expect, like video decoding and encoding, etc.

Yet another AMD fan who cannot read. I said there are critical issues and missing features in the RX 5000 series and you again try to counter that ... with an earlier generations AMD product. Open Source/AMD fans seem not to possess any argumentation skills.

Quality open source drivers, my ass.

[Raka555]

BSD failed relative to Linux. It didn't get the same level of attention, and the vast majority of software available was built with other OSes in mind.
Of course, success here is a matter of perspective, but to say BSD didn't fail would also mean that ReactOS or Haiku didn't fail.

In the same way that Wayland failed relative to X-Windows ?

[Raka555]

If people think linux failed on the desktop now, wait till NVidia gives them a solid middle finger ...

I have to agree with Birdie on this one. I feel he's pain.
I myself went through the same crap with an AMD card many years ago. You buy it because of all the hype and support around the opensource drivers. And the Phoronix benchmarks look good.
Only to discover, after you bought the card, all the bugs and issues
At one point you just get tired of all the struggling and just buy the equivalent NVidia card. Instant bliss. Everything just works...

The much hated Intel is actually the real heroes here.
They always have excellent drivers for their GPUs. I never experienced any issue in 15 years of using the Intel drivers. All opensource.
I hope their discrete GPU will be a success and deliver playable performance for most games.