Rust-Written Apple DRM Linux Kernel Driver Renders First Cube

[qarium]

Linus Torvalds is using an m2 Macbook Air to build aarch64 kernels on. And I know developers who using these systems for the same purpose. Some distros like CentOS don't allow you to cross-compile (official) packages -- which means official packages have to be compiled on a native arch. So M1/M2 systems are really attractive to (some) developers and package maintainers for that reason .....but there's a million other reasons. The performance/power ratio, build quality, and aesthetics are pretty damn good -- and that's enough to attract more mainstream users.
Bluetooth has been working for months and yes there's a 2.4Ghz wifi/bluetooth co-existence issue that's currently being worked on, but I think you're digressing here a bit. There's a difference between being fully-functional and being usable. You were mocking people for celebrating a spinning cube. But that spinning cube represents progress -- as in this project is progressing in the direction of becoming fully-functional. I think most people could get by without a webcam. Although let's be honest, even when/if these systems become fully supported by Linux -- you'll just find something else to complain about....because that's your thing.
I'm pretty sure aarch64 Linux users represent less than %.1 percent of the entire gaming market. I just don't get who would be buying an M1/M2 system for gaming on Linux. That sounds ridiculous.
So let me get this straight. You're saying that the general population of all computer users would care more about what frame rates they're getting with Doom then they would about having a stable desktop experience. Again...I don't really know what to say except for that's pretty damn insane.

to unterstand him is very simple he want to game on his notebook and right now with x86 on rosetta2 and directX translated to metal the apple m1/m2 are bad in gaming...

this is his only reason to hate the apple m1/m2...

apple could fix this by license RDNA3 design with FSR3.0 and Raytracing hardware... and also opensource drivers.

but right now no one buys a apple m1/m2 to game ...

[qarium]

also no one ever did try a game with native 64bit ARM binary with native metal/WebGPU support with FSR2.1 ...

with that the result could be the apple m1/m2 is good for gaming.

but these intel/Nvidia/microsoft monopole lovers will never accept such a benchmark only if x86 is translated to ARM ISA and DirectX is translated to metall and the windows game is translated by proton then they accept the result. (to make sure apple lose the benchmark)

[Dukenukemx]

Linus Torvalds is using an m2 Macbook Air to build aarch64 kernels on. And I know developers who using these systems for the same purpose. Some distros like CentOS don't allow you to cross-compile (official) packages -- which means official packages have to be compiled on a native arch. So M1/M2 systems are really attractive to (some) developers and package maintainers for that reason .....but there's a million other reasons. The performance/power ratio, build quality, and aesthetics are pretty damn good -- and that's enough to attract more mainstream users.

Aesthetics? Now I know you're an Apple owner.

Bluetooth has been working for months and yes there's a 2.4Ghz wifi/bluetooth co-existence issue that's currently being worked on, but I think you're digressing here a bit. There's a difference between being fully-functional and being usable.

I'm sorry I'm too busy using my bluetooth headset and watching a YouTube video at the same time. You know, a function most people you described would use. Did you have a point?

You were mocking people for celebrating a spinning cube.

I'm mocking people for not holding Apple accountable for the lack of progress.

But that spinning cube represents progress -- as in this project is progressing in the direction of becoming fully-functional.

My experience with this says that goal will never be reached. I use Linux for gaming and I'm not at 100% functionality. I have to make sacrifices that I'd like to see go away.

I think most people could get by without a webcam.

I think you're lowering the bar low enough to excuse this.

Although let's be honest, even when/if these systems become fully supported by Linux -- you'll just find something else to complain about....because that's your thing.

Let's be honest, you're doing the same in reverse. The only difference is that I'm correct.

I'm pretty sure aarch64 Linux users represent less than %.1 percent of the entire gaming market. I just don't get who would be buying an M1/M2 system for gaming on Linux. That sounds ridiculous.

Baby steps first. You can play Doom, Quake, and even Doom 3 on OpenGL 2.1 since there's so many open source projects that have ARM support. You can include emulators as well. You can get away with a lot without needing to worry about native games.

So let me get this straight. You're saying that the general population of all computer users would care more about what frame rates they're getting with Doom then they would about having a stable desktop experience. Again...I don't really know what to say except for that's pretty damn insane.

Nobody said that was a choice. You can have your stability and eat your gaming as well. My point is that Apple could make this process go a lot faster if they just got a little involved. This is to others benefits, not mine because I'll never make the mistake of buying an Apple product. Get Apple involved and we can go from getting basic features working to bug fixing and performance improvements. At this rate you're years away from 100% functionality in Linux. Stop lowering your standards and make Apple donate code or hire people to help.

[akira128]

Stop lowering your standards and make Apple donate code or hire people to help.

Soooo
Step 1: Hold out your hand and politely ask Apple to develop Linux drivers.
Step 2: Don't do a damn thing until Apple complies with step 1

It's such a sign of the times that your grand plan involves NOT DOING A DAMN THING.
Previous generations would have scoffed at that sort of highly-entitled behavior.
Go back to watching porn and playing games dude -- mission accomplished right?

[qarium]

Soooo
Step 1: Hold out your hand and politely ask Apple to develop Linux drivers.
Step 2: Don't do a damn thing until Apple complies with step 1

It's such a sign of the times that your grand plan involves NOT DOING A DAMN THING.
Previous generations would have scoffed at that sort of highly-entitled behavior.
Go back to watching porn and playing games dude -- mission accomplished right?

even if apple would provide the perfect linux driver in the limitations of the hardware they have he would not buy it.

the reason is simple: he want to game on the notebook he buys and the apple gpu in M1/M2 lags hardware features to make energy efficient gaming possible.

he will go and buy a intel/AMD with RTX4000 Nvidia GPU notebook with DLSS3.0 with FAKE FRAMES result in 25% lower power consumption.

and also all his games are "x86" and DirectX or Vulkan and no metal or WebGPU ...

if you need to translate x86 to ARM with Rosetta2 and you need to translate vulkan/DX12 to metal and no GPU with RayTracing hardware acceleration and no FSR3.0/DLSS2./3.0 upsampling hardware acceleration ...

with conditions like this he will never buy a apple notebook not even if apple provide perfect linux drivers.

so now is the big question why he even want linux driver if in the end he does not buy the apple product anyway.

thats the big question... he always talks and in the end talk is cheap and in the end he does not buy apple laptop anyway.

[tunnelblick]

Incredible how one guy with uninformed and made up things was able to, again, derail the whole thread. I wish there was a block function on this forum.

[qarium]

Incredible how one guy with uninformed and made up things was able to, again, derail the whole thread. I wish there was a block function on this forum.

its not the case that Dukenukemx is uninformed he is informed but he accept only a very selective selection of tests as relevant,

he made up his oppinion on only 2 facts: some 6nm AMD notebooks have similar performance in multicore and multicore performance per watt. and the second one: he watches tests of windows games with x86 with directX binary translated to macos and 'ARM with rossetta2 and translated to metal... and this is for him the proof that apple M1/2 is bad.

he does not test native 64bit ARM binary on Metal/WebGPU with FSR2.1 .... of course not.

people try to educate him that single core performance is more relevant (even for games) and that most tasks who M1/M2 loses general compute benchmarks these benchmarks are irrelevant because M1/M2 gas ASICs to perform the tasks what does a video codex benchmark matters on the ARM cores if apple hardware has ASIC video encode unit who performs the task much better,

so in my point of view the world view of Dukenukemx is complete idiotic.

[tunnelblick]

its not the case that Dukenukemx is uninformed he is informed but he accept only a very selective selection of tests as relevant,

If you say "I think they recently fixed Bluetooth but you can't use it with WiFi at the same time." then this is bending the truth to fit your narrative. It is intentionally misleading and not telling the truth IF he is informed.

so in my point of view the world view of Dukenukemx is complete idiotic.

I cannot disagree with that, sadly. But that is already giving the user more attention than is deserved here.

[Volta]

for the performance you get they are not expensive. They are rather cheap actually. Also.. hardware isn't that easy to brick anymore.

For the crapOS - being an only usable option currently, they're far too expensive.

[piotrj3]

its not the case that Dukenukemx is uninformed he is informed but he accept only a very selective selection of tests as relevant,

he made up his oppinion on only 2 facts: some 6nm AMD notebooks have similar performance in multicore and multicore performance per watt. and the second one: he watches tests of windows games with x86 with directX binary translated to macos and 'ARM with rossetta2 and translated to metal... and this is for him the proof that apple M1/2 is bad.

he does not test native 64bit ARM binary on Metal/WebGPU with FSR2.1 .... of course not.

people try to educate him that single core performance is more relevant (even for games) and that most tasks who M1/M2 loses general compute benchmarks these benchmarks are irrelevant because M1/M2 gas ASICs to perform the tasks what does a video codex benchmark matters on the ARM cores if apple hardware has ASIC video encode unit who performs the task much better,

so in my point of view the world view of Dukenukemx is complete idiotic.

You currently have hardware encoders on all Intel, AMD, and Nvidia GPUs. Most of them are far more advanced and produce higher quality output then Apple does at same bitrate. Intel already has AV1 encode support and Nvidia rtx 4000 as well.

Also no, hardware encoders are still behind software encoders, just hardware encoders are simply a lot faster (but their quality is behind and in case of Apple A LOT behind).