Live data from Hacker News

Same-day upstream Linux support for Snapdragon 8 Elite Gen 5

qualcomm.com

221–230 of 249 posts

Re: Same-day upstream Linux support for Snapdragon 8 Elite Gen 5

#221
post #204

Earlier quoted context omitted.

The very first they got wrong is how hard it is to buy Dell XPS with Linux on most European countries. Naturally nowhere to be found on PC stores, and online I never found it on sale in the Dell store.

Why would you want to do that anyway?

Because that is how normies buy computers, duh?

Go into store, get to talk with a "genie", whatever upsells to meet their KPIs, and leave with another computer that counts to the world market share in desktop units.

Which for all practical purposes means Windows, macOS, Chromebooks, iPad Pro or Android tablets with keyboard for Dex, HyperOS, HarmonyOS NEXT , and zero GNU/Linux devices.

Re: Same-day upstream Linux support for Snapdragon 8 Elite Gen 5

#222
post #82

Earlier quoted context omitted.

I worked at Linaro, who was contracting for Qualcomm. Qualcomm were pushing for some protected hypervisor called Gunyah (which had its own Linux interface and needed a new qemu port) that apparently no one liked. I tried to port it to KVM [1], but upstream folks (mostly Google) outright rejected the port. Otherwise KVM would have been available on QCOM boards. You can still try it. I have a Linux kernel and a Qemu po…

Upstream would accept a patchset that exposed an independent Gunyah-specific UAPI (why not the same one as downstream — crosvm already supports that) instead of pretending to be KVM (it's not a "port", you can't port a hypervisor to a hypervisor). KVM is available on current compute platforms (laptops) if you escape to EL2 via slbounce; and on Glymur (X2E) it will be available by default (yay!).

That's not how operating systems work. KVM is both an interface and a hypervisor. Just as we have different hypervisor implementations for amd, intel, arm and others all abstracted behind the same KVM interface, there is no reason the same can't be done for Gunyah. Userspace does not have to know anything about that. KVM already supports svm and vmx for amd and intel on x86. Why is something similar can't be done for Arm? Plus now there is pKVM.

I just don't understand this argument of a separate interface. The only reason you want to do that is to decouple from the KVM community, but that introduces a shit tone of duplicated effort and needless fragmentation to the virtualisation software ecosystem hindering your users from enjoying the existing upstream tools they already know about. In other terms, vendor locking and shitty downstream experience.

Re: Same-day upstream Linux support for Snapdragon 8 Elite Gen 5

#223
post #203

Earlier quoted context omitted.

Usually GPU vendors (Nvidia, Intel, AMD) provide a way to download and install drivers manually (on Windows), including specific versions or older versions. Qualcomm is an outlier in this case.

Note that those drivers usually only work well in desktops, on laptops the GPU might have gone through OEM adaptations on the motherboard integration, and a driver from GPU vendors might have issues. A common example is overheating, because the way the OEM has done their device isn't a setup that the driver knows about. Which is why on laptops, the drivers if available have to be from the OEM themselves.

It might be more fair to say that there's simply no standard way to do power management for discrete GPUs in a laptop, or to integrate such power management with Windows and whatever power management it's trying to do. And the lack of a clear "right" way to do things means laptop OEMs use this area for product differentiation with their own shitty special sauce software and firmware hacks.

If installing drivers that come directly from NVIDIA onto a laptop can cause that laptop's GPU to overheat in the sense of getting so hot it fails to function properly or has degraded reliability, that's entirely NVIDIA's fault. If by "overheating" you just mean drawing more power and causing the fans to get louder than they would in an out of the box configuration, the blame for that should be shared between NVIDIA, Microsoft, and the laptop OEM, but you shouldn't blame the user for doing something that should work and would work if those three vendors could cooperate.

Re: Same-day upstream Linux support for Snapdragon 8 Elite Gen 5

#224
post #86

Earlier quoted context omitted.

Get a DGX spark. Ships with aarch64 Ubuntu 24.04. Tons of cores and RAM. Very quiet and small UEFI bootloader - I installed Ubuntu 25.10 and ESXi arm edition just by booting the ISO usb-c power input (kinda cool) Insane connectx 200GbE RoCE networking 10GbE Ethernet Oh and an nvidia gpu with cuda and access to 128GB of unified memory It would be perfect if it had some kind of BMC or IPMI/redfish and an exposed PCIE s…

This might sound silly question, but those of you who have digits/spark machine, has anyone run Fedora on it? I kind of ran away from Ubuntu back to Fedora because reasons. Bonus question, far-fetched, steam and games with FEX?

Steam gaming with FEX is possible on the DGX Spark. The GPU is approximately a mobile 5070 with much less memory bandwidth. The CPU cores are relatively weak, especially after the instruction set translation overhead. There's a lot of stuff that's playable, but the performance is laughably bad for a $4000 machine.

Re: Same-day upstream Linux support for Snapdragon 8 Elite Gen 5

#225

Earlier quoted context omitted.

Those that are implemented have been very reliable in my experience, I think that labeling them “alpha-quality by Linux standards” is a ridiculous claim

Then you need an Intel or AMD laptop as a frame of reference. M1 is implemented as-is with much of the silicon's onboard accelerators entirely dark. Hardware accelerated video encode/decode is a lost cause, Thunderbolt will likely never happen, NEON is your fastest SIMD accelerator and cpuidle is still not really figured out. Those are all perfectly acceptable limitations for a POC. And the GPU drivers are particular…

There's a big difference between saying that many of the drivers are missing entirely, and saying that the drivers that exist are alpha-quality.

Re: Same-day upstream Linux support for Snapdragon 8 Elite Gen 5

#226

Earlier quoted context omitted.

Apple is a non-sequitur. Tuuxedo is giving up on Qualcomm in favor of AMD and Intel. Do AMD and Intel require Tuuxedo to pay for premium support in order to get working Linux drivers? No, of course not. Qualcomm's support for Linux is embarrassing when you compare it to pretty much any processor manufacturer except Apple.

Apple doesn't actively lock and act hostile towards the reverse engineering folks. They just do their thing and sometimes do small jests to allow these folks to boot anything they want and play with the hardware. ...and while they're standing away from GPL stuff, they do have a dedicated site for Open Source software: https://opensource.apple.com/ Go to releases, and see what they do there.

if you are in the business of buying parts, assembling them, and selling the assembly under your brand (as tuxedo and others like them are), then Apple is a indeed not an option. Their chips might be the best, but you still can only get them in Apple devices, in the specs Apple provides and with no official support for any OS beside macOS / iOS.

Re: Same-day upstream Linux support for Snapdragon 8 Elite Gen 5

#227

Earlier quoted context omitted.

Thanks for the constructive framing. My gut reaction was "great job, you're doing the minimum"

If I was to guess, id say someone did this as a side project at QC because maybe they like FOSS and want to give something back. Given the partnership between QC and MSFT for laptops and Google for android, I won't be surprised if we never see interest from QC for any real linux hardware.

They hired Linaro (Linux Consultancy) for it so not just a side project

Re: Same-day upstream Linux support for Snapdragon 8 Elite Gen 5

#228
post #57

Earlier quoted context omitted.

Ooh, thank you for this, I might try it on my m4 mac. Any tips or anything I should be aware of?

i used the UTM app from the App Store, and when creating a new instance, i select the Linux icon, which exposes the selection to enable Apple Silicon hypervisor rather than Qemu. it works perfectly. and it's fast. just great. I had used Asahi before, dual booting, which was a pain in the neck. this meanwhile is perfect.

Oh wow very simple, thank you

Re: Same-day upstream Linux support for Snapdragon 8 Elite Gen 5

#229
I was going to write a snarky comment, but in the "if any of Qualcomm leadership is listening" I'm going ask a question:

Why is any of this needed when the kernel is full of platforms that are forward compatible with the Linux kernel and boot and generally operate on day one, without a huge pile of changes?

What does it benefit the user to have a huge pile of proprietary implementations of devices they frankly don't care about? Ex: just about anything related to power management? Why can't QC adhere to industry standards when they implement standard devices, ex: USB? Why can't these platforms adhere to industry standard firmware interfaces rather than custom mailbox interfaces?

Re: Same-day upstream Linux support for Snapdragon 8 Elite Gen 5

#230
post #35

Woah, this is amazing. I’ve been looking for an ARM Linux machine for a while and ended up about to get M2 Pros in a rack running Asahi. It has been near impossible to get a Snapdragon Elite machine. The IdeaCentre or whatever is 2x the cost / performance and as far as I know is poorly supported. This changes the game. I’d rather use native Linux than Asahi (though the latter is amazing).

How is Asahi not native?

Asahi is also still a platform with a huge pile of out of tree patches on top because the platform itself is pretty unusual, requiring for example, a 16K page size kernel which is unlike pretty much every other arm Linux platform.
Post reply on HN