AMD Is Exploring A Very Interesting, More-Open Linux Driver Strategy

[Luke_Wolf]

Never is an extremely long time! With the improvements to C++ and the advancements made compiler wise I can see future acceptance of C++ in portions of the kernel code base.

It has nothing to do with technical reasons it has everything to do with Linus hates C++ and OOP and is too stuck on C for his own good.

This:

http://warp.povusers.org/OpenLetters/ResponseToTorvalds.html

is well worth reading

[cutterjohn]

I read about this first on /. this morning, and my first thought is how is this going to mesh up with their windows driver development as ATM BOTH nvidia and ati share what they describe as a "majority" of their driver code between windows and linux(and whatever else they support from time-to-time)...

Really sounds to me like if they're not careful the linux drivers(from a catalyst) perspective are likely to end up suffering even more unless they can also leverage that into windows drivers.

Might be helpful overall as catalyst is still crappy compared to nvidia, where everything has just worked for aeons now excepting for the very rare borks. Case in point when I build my "pure" AMD systems recently I finally found out that under windows ati finally discovered how to unload their old vid driver then reload the new w/o having to reboot first. i.e. IME w/4850 when it was supported was that it unloaded the windows driver then left me with a blank screen, so I played the guessing game of "is it done yet?" then eventually rebooted, which was quite a shock to me compared with my nvidia experience, which was unload/reload I have display again and could SEE if it wanted to reboot AND WHEN it was finished...

Seriously though ati REALLY needs to get their act together on vid playback as even the R9 280X STILL tears like mad in BOTH windows AND linux regardless of the settings... Trues it's better than 4850 days BUT it's still MUCH worse than nvidia EVER was...(that was another shock of the 4850, i.e. how crap vid playback was for what SHOULD have been a VERY capable card...)

[Alejandro Nova]

I am not quite sure how they would do this:
- Mantle on Linux? Sorry, resources are scarce, maybe we can later look at if it is even worth our time.
- UVD for RS780/880/RV790? Sorry, resources are scarce, maybe later, maybe never.
But:
- Major changes in the driver to be able to use the radeon kernel module, which will take a significant amount of developer time? Hey, we totally can do that!

WTF?

- Major changes in the driver to be able to use the radeon kernel module, which will take a significant amount of developer time? Resources are scarce, so, we'll maintain only one set of kernel interfaces instead of two, reduce our maintenance burden, and free up developers to even think about Mantle.

UVD for those cards? Life is hard.

[Alejandro Nova]

I read about this first on /. this morning, and my first thought is how is this going to mesh up with their windows driver development as ATM BOTH nvidia and ati share what they describe as a "majority" of their driver code between windows and linux(and whatever else they support from time-to-time)...

Really sounds to me like if they're not careful the linux drivers(from a catalyst) perspective are likely to end up suffering even more unless they can also leverage that into windows drivers.

Might be helpful overall as catalyst is still crappy compared to nvidia, where everything has just worked for aeons now excepting for the very rare borks. Case in point when I build my "pure" AMD systems recently I finally found out that under windows ati finally discovered how to unload their old vid driver then reload the new w/o having to reboot first. i.e. IME w/4850 when it was supported was that it unloaded the windows driver then left me with a blank screen, so I played the guessing game of "is it done yet?" then eventually rebooted, which was quite a shock to me compared with my nvidia experience, which was unload/reload I have display again and could SEE if it wanted to reboot AND WHEN it was finished...

Seriously though ati REALLY needs to get their act together on vid playback as even the R9 280X STILL tears like mad in BOTH windows AND linux regardless of the settings... Trues it's better than 4850 days BUT it's still MUCH worse than nvidia EVER was...(that was another shock of the 4850, i.e. how crap vid playback was for what SHOULD have been a VERY capable card...)

AFAIK the Windows Catalyst drivers are almost as crappy as Linux ones, but Microsoft has two things Linux won't have: 1. a "stable kernel API", because the Windows kernel is feature-frozen since 10+ years, and 2. a tremendous Q&A department to introduce as many workarounds as necessary.

[_SXX_]

2. a tremendous Q&A department to introduce as many workarounds as necessary.

Sure, this why built-in drivers of Windows 8/8.1 for AMD legacacy graphics have no OpenGL support and they can't be replaced even by AMD installer.

[Nille]

Case in point when I build my "pure" AMD systems recently I finally found out that under windows ati finally discovered how to unload their old vid driver then reload the new w/o having to reboot first. i.e. IME w/4850 when it was supported was that it unloaded the windows driver then left me with a blank screen,

This works for me since Vista and my old HD3850

[_SXX_]

Though AMD mobile drivers for Windows are so bad that their support website contain zero direct links to them and forums about laptops a full with topics "update AMD drivers on laptop XYZ". Some models of laptops with dual graphics (HD4250 integrated + 5XXX discrete) don't get official driver updates at all and you have to install "custom modified" drivers to some web forum to make this hardware work properly.

Linux users at least have open source drivers that might be not suitable for games, but at least works properly and for any other purpose. So yeah I agree that for some part of users AMD drivers on Windows way forse that Linux one.

[GreatEmerald]

I read about this first on /. this morning, and my first thought is how is this going to mesh up with their windows driver development as ATM BOTH nvidia and ati share what they describe as a "majority" of their driver code between windows and linux(and whatever else they support from time-to-time)...

The "majority" is not in the kernel-facing code (the Linux and Windows kernels are nothing like each other, remember?).

Seriously though ati REALLY needs to get their act together on vid playback as even the R9 280X STILL tears like mad in BOTH windows AND linux regardless of the settings... Trues it's better than 4850 days BUT it's still MUCH worse than nvidia EVER was...(that was another shock of the 4850, i.e. how crap vid playback was for what SHOULD have been a VERY capable card...)

Yeah, about that... My GTX 660 is tearing like crazy, no matter the options set. In fact it's tearing this very page when I merely scroll. So no, it's not worse than NVIDIA in the slightest.

[smitty3268]

I don't get that. They still would have to collaborate with the radeon developers to get release-day support for new hardware, tehey still have to get all the features work that aren't currently supported with radeon, so where exactly would they reduce the time? This sounds like: Hey, we offload all this work to the FOSS developers which will reduce the work we have to do and as a nice side effect they can't blame us when new hardware doesn't work, we just say "Here is the documentation, why don't you just implement it?"

It's not that difficult to understand. You know how every time a new kernel comes out, people complain that the driver doesn't work anymore? And then they wait a few months for AMD to patch the driver before it works? None of that would be necessary anymore. The reason behind that is the entire reasoning behind putting open source drivers in the kernel instead of maintaining them out of tree. Google it if you don't understand, there are thousands of posts out there which can explain it.

Sorry, but AMD has a nice track record of screwing their customers and to me this sounds awfully like another attempt.

It sounds like you've already made up your mind and will never see anything they do in any other light. Well, that's your right, but it doesn't mean you are correct.

[xeekei]

Yeah, tearing vs stuttering, the "choose your poison" of the modern graphics experience. I hope Gsync takes off, and won't be Nvidia exclusive, to fix that one.