Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Initial M1 support merged into Linux SoC tree

git.kernel.org

Post reply on HN
Print view View original

291–300 of 337 posts

Re: Initial M1 support merged into Linux SoC tree

#291

Post by rowanG077 »

rvz wrote: post #152
stepbeek wrote: post #88

Will this have an impact of tools that depend on linux like docker on mac? Maybe android emulation? If so, then I'm super excited for it. I'm using an m1 macbook air right now and it's the first compelling mac laptop for me since 2015. It's absolutely amazing. But for any workload that uses docker, my linux desktop is streets ahead.

Last time I checked, the Android Emulator running on Apple Silicon still has issues and isn't functionally as good as running it on an Intel Mac. There is no point in getting the M1 Macbook (which is last year's model) since they will either announce a M1X or even a M2 Macbook this year which will most certainly be more performant and optimised than the last generation. One of the several reasons why I stayed away fr…

And once the M2 is released there is no point in getting it anymore since then it's the outdated model and the next one will be announced soon ;).

Re: Initial M1 support merged into Linux SoC tree

#292

Post by vetinari »

rkangel wrote: post #190
piokoch wrote: post #77

Earlier quoted context omitted.

I was investigating recently options which laptop I can buy to be able to use Linux without any issues. I've checked Lenovo Legion 5 with AMD processor. Looks like good machine for a price, but touchpad is not gonna work out of the box on the latest stable Ubuntu. Kernel upgrade is needed (or PopOs distro that already has needed patch). For me kernel update does not sound as something big, at least I know that someth…

If you want to use Linux on laptops, the best option is Fedora. Of all the mainstream dependable Linux distros it has the most up to date Kernel and other bits (like Gnome). This means that its support for hardware is usually the best. On Lenovo laptops with Intel graphics everything works perfectly out of the box with no issues (all the way down to fingerprint readers etc.). I also find it to be stable and generally…

It helps, that some Thinkpads do ship with Fedora OOB. Lenovo worked with Redhat to have the support for these models upstreamed to the respective projects.

Re: Initial M1 support merged into Linux SoC tree

#293

Post by DCKing »

533474 wrote: post #70

Why do we still buy macs? Apple can pull the plug anytime. Kudos yo the huge undertaking but I personally think we wouldn't need this type of gargatuan effort if we just refused to buy macs without linux support. Why support such a closed system?, enlighten me...

This goes for every hardware platform doesn't it? The Raspberry Pi foundation could decide to have Raspberry Pi systems boot only Windows 10 IOT only starting tomorrow. But they don't have any interest to do so.

Apple went out of their way to add support for third party OSes to Macs. They actively made tools to allow the Asahi Linux folks and others to boot unsigned OSes on M1 Macs. They would have had much less work - both in terms of making and maintaining the tools as well as customer support work - if they kept the boot system more similar to the iPad boot system it derives from. In general really Apple doesn't seem to be interested in closing down the Mac at this level. Hence why you can (attempt to) run third party OSes on virtually all Macintoshes going back to 1987 [1].

I for one am happy with that. If I buy an Apple Silicon Mac, I won't be running Linux on it - at least not on bare metal. But I might do that in 10 years after Apple's support has ended.

[1]: NetBSD claims to still support the Mac II, a Motorola 68030 based Mac release in March 1987: https://wiki.netbsd.org/ports/mac68k/

Re: Initial M1 support merged into Linux SoC tree

#294

Post by the_french »

anthk wrote: post #276
the_french wrote: post #224

Earlier quoted context omitted.

On some fronts it even feels like things have regressed . Trying to resize an encrypted partition is way, way too difficult. And it seems like GParted doesn't handle LUKS so you're back to manually typing block offsets on the command line. If you wanted to use a TPM to store the FDE passphrase, well, you have the patience of a saint. Compare this to Mac or Windows where you click a single button and it's all setup, i…

Get OpenBSD. FDE is dumb easy. I am not kidding.

from the cursory read of the documentation on the openbsd site it doesn't seem to support TPMs. Setting up the encryption of a disk paritition does seem easier though.

Re: Initial M1 support merged into Linux SoC tree

#295

Post by segfaultbuserr »

redm wrote: post #211

After reading articles last year on Linus insinuating it would be "unlikely" that Linux would appear on the M1 [1] [2], I drank the cool-aid and took that at face value. I guess I should have known better, just like nature, open source "finds a way". [1] https://www.realworldtech.com/forum/?threadid=196533&curpost... [2] https://news.ycombinator.com/item?id=25216680 etc..

> After reading articles last year on Linus insinuating it would be "unlikely" that Linux would appear on the M1 [...] just like nature, open source "finds a way".

Not necessarily.

Reverse engineering the unknown architecture and porting the Linux kernel to the new machine is impressive work that we should support and celebrate.

But making it to be a usable desktop or laptop computer (e.g. with functional power management, 3D acceleration and other drivers) is another matter. See Nvidia graphics. Even after 10 years, some cards are still not fully supported by Nouveau due to the lack of documentation. I'm afraid that Apple M1 is going to stuck at a similar state for a long time - technically, Linux boots, but it's not a machine you'd like to browse the web, write code, or carry around.

Re: Initial M1 support merged into Linux SoC tree

#296

Post by mrtranscendence »

hobonate wrote: post #231
tgv wrote: post #202

Earlier quoted context omitted.

I've written this before, but it's not going to happen without a large amount of apps, that people actually want and can use. And that's not going to happen unless there's all the necessary tooling, especially for UI. mac/iOS and Windows have great support for building user-oriented applications. And then it still needs a centralized way to monetize the development work. And to get all that, you need a ton of paying…

There's already GTK and Qt. What more do you want? > And then it still needs a centralized way to monetize the development work. I disagree: once you put money into the equation, the quality of packages will go down as programmers try to shake as much money out of the users. Compare programs like MATLAB vs GNU Octave or Photoshop vs GIMP for example. > And to get all that, you need a ton of paying users, not a bunch…

You have a skewed view of what it’s like on Windows or macOS. I certainly haven’t paid for a bunch of crappy, trivial apps — the utilities I use are all free (and don’t upload my data to the cloud, nor make me watch ads).

And are you saying that gimp is a superior product to photoshop? It’s not like you can’t get gimp or octave on non-free OSes, but for some (good) reasons illustrators and scientists and engineers choose to pay for photoshop and matlab.

Re: Initial M1 support merged into Linux SoC tree

#297

Post by mjg59 »

jcelerier wrote: post #196
mjg59 wrote: post #140

Earlier quoted context omitted.

As mentioned by AnssiH, this hasn't been true for a long time - https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/lin... was merged in 2007. I wrote up the reasoning back in 2008 at https://mjg59.livejournal.com/85923.html , but since then Linux should always have been behaving identically to the most recent version of Windows as far as OSI responses go.

most features of my laptop don't work reliably (MSI GS65) if I don't use the acpi_osi line, why would that be in that case ? I'm on archlinux with pretty much always the latest kernel

What argument are you providing? In some cases if you pretend to be an older version of Windows then things will work better, since it's possible for the kernel to end up pretending to be a new version of Windows without all the relevant semantic changes in the drivers having been made.

Re: Initial M1 support merged into Linux SoC tree

#298

Post by kitsunesoba »

RantyDave wrote: post #96
ginko wrote: post #80

Earlier quoted context omitted.

Anecdotal I know, but I very recently got a macbook air after years of practically exclusively using Linux and the opinionated user interface drives me crazy. Here's a few pain points: - Scroll direction can only be set globally so either touchpad or mouse scrolling feels awkward for me - You need an extra click on another window before you can interact with it - No focus-follows-mouse (there used to be 3rd party plu…

There's a certain Zen to using Apple interfaces (IMHO): You have to relax and just accept that this is how it is, and it's best to get used to it. I fully understand if you find this to be not really good enough, but this is the only way to make them work for you.

There's truth to this. After years of using macOS as my primary OS, both Windows and Linux w/various DEs become progressively more irritating the more windows are open because they both expect micromanagement, be it from the user, or automated as with a tiling WM.

Under macOS, windows just kind of exist as they want to be and pile up until you intentionally change them, and the environment doesn't push you to manage it, so you end up not putting much thought into anything but the 1-3 windows you're using at that exact moment.

The mac approach taps a bit into spatial memory too, which makes auto-tiling like with i3 irritating because then windows no longer have an inherent size or position — it all continually changes as windows open and close unless I start maintaining window sets, so at the end of the day I'm still putting more thought into management than I would've under macOS.

Because of this it's long been a desire of mine to have a Linux DE that closely mimics macOS. Some will point to GNOME, but it's more like iPadOS if it were adapted to the desktop. Elementary gets closer but still misses the mark. It would be a neat hobby project to take up but figuring out how to get started is rough, especially with all Wayland example WMs and tutorials being for tiling instead of traditional floating.

Re: Initial M1 support merged into Linux SoC tree

#299

Post by fulafel »

nicoburns wrote: post #266
fulafel wrote: post #155

Earlier quoted context omitted.

The secret to the lightweight nature of Docker is that it just uses various namespacing, cgroup etc functionality provided by Linux with no virtualized OS kernel or hw virtualization mechanisms involved. (Yep you need ARM Docker images when running on ARM of course)

It does on linux. On macOS and windows it does still do this... but in a linux VM.

Yep. Linux gaining HW support for the M1, and Docker on your penguinized mac, was the context I thought.

Re: Initial M1 support merged into Linux SoC tree

#300

Post by comex »

tyingq wrote: post #243
sillysaurusx wrote: post #125

Earlier quoted context omitted.

(Side note: I'm happy to see that shilling your patreon is becoming a socially accepted norm. Best of luck with your work!) One thing I was surprised by: LuaJIT doesn't compile on an M1. LuaJIT 2.1.0 technically compiles, but it core dumps whenever you have a syntax error, which makes it quite a bit less useful. A reason for this might be due to the lack of __threadlocal support in M1's (apparently, or so says clang)…

>lack of __threadlocal support in M1 An informative post about this from the Zig folks: http://www.jakubkonka.com/2021/01/21/llvm-tls-apple-silicon....

That's not a lack of support, it's just LLD, the LLVM linker, being broken. I hear LLD for Mach-O is being worked on again, after a long period of stagnation where it was sort of but not fully working. But the default linker on Apple platforms is still ld64.
Post reply on HN