Corellium Posts Very Early Linux Port To Apple M1 Macs

[WorBlux]

How is it a waste of time to make Linux available on it? Isn't being able to run Linux almost anywhere one of the things the Linux community is typically proud of?

Because the chip designer/vendor has offered exactly zero support of documentation, and could lock you out of all these devices with a simple firmware update. The only assurance you have is a weak public statement that they won't interfere, but no real contract, dev kit.... etc that would give anyone any standing to sue should they buy the device to run linux and later be locked out (see the PS3 fiasco where there was a lawsuit required for compensation that only gave them 10% or the purchase price back). And Apple isn't advertising it as a feature anywhere.

They don't make money from their hardware, they make money from their walled garden, they have little reason to help any alternative OS. At best these developers are just spending thier time and attention on a project to build an anti-trust defense for one of the richest companies on earth, and that has almost zero transfer of utility outside of apple systems, and that could be cut off at any time should it actually threaten apple profitability.

I can understand Coreillium wanting to do it and paying thier own devs to do so. Even if they do get cut off later, understanding the silicon better helps their core business model. But AsahiLinux as a community project trying to bring a "polished linux experience". I don't really get that.

[GruenSein]

waste of time and talent

First of all, it is up to everyone to decide if spending time on something is worth it. If whoever is working on this wants to run Linux on their M1 Macs it is worth it.

Secondly, I respectfully disagree with you in a general sense, too. The M1 is arguably the best currently available ARM SoC targeted at customers. The M1 Macbooks are nice machines. How is it a waste of time to make Linux available on it? Isn't being able to run Linux almost anywhere one of the things the Linux community is typically proud of? Why not use this machine in the future to have a light yet powerful laptop with a decent screen and lots of battery life for your Linux needs? Most mainboard vendors don't help supporting coreboot either. Nonetheless, they make decent mainboards on which people would like to run free software. Is coreboot a waste of time?

[Aryma]

waste of time and talent

[OneTimeShot]

They've already distributed a "very early beta" build.

Also distributing broadcom binary blobs, and possibly a custom busybox build

I note that you linked to Twitter, not to binaries. Any reason for that? Let's guess...

[OneTimeShot]

So you did not even read the title of the article that your posting comments for?

yes - and I read the blogs, and took a quick look at the source code on their GitHub site? What's your input?

Maybe you can point me to a location where they are distributing the binary? Or did you not click on that?

[Etherman]

kernel 5.4?

[curfew]

It's only GPL violation if they distribute it. You are free to mess around with GPL code on you own machines however you want.

So you did not even read the title of the article that your posting comments for?

[Awesomeness]

They've already distributed a "very early beta" build.

Also distributing broadcom binary blobs, and possibly a custom busybox build

The GPL doesn't say that both binary builds and source archive must be available at the same time. In fact a written offer to send the code via mail (with media and shipping cost to be paid by the user) is enough. There can't be a profit attached to ending over the code but it's still a hurdle one must take to get the code and it's all fine with the FSF.

[waxhead]

I am not fan of Apple , but the M1 seems interesting. Just maybe there is a faint chance that the M1 can kick off a RISC-V based CPU soon that will have a hybrid of energy efficient cores and high performance cores. Of course that will be interesting (and challenging) for the process scheduler as it can run processes that chew on large amounts of data on the efficient cores on perhaps a slower scheduling interval, or if you nice the process it perhaps makes sense to run it on energy efficient core(s) chugging along in the background.

Not sure how the CFS on Linux handle hybrid configurations , but Intel is lagging behind these days which is good for making some healthy competition.

[aspen]

i think competition between Corellium and AsahiLinux can be interesting and that the first team that gets merged upstream "wins".

From my point of view it is interesting that Corellium has something working, but how AsahiLinux is approaching the project documenting every single step, making clear statements on how they'll RE, open to collaborate with whoever that wants to follow the strict but necessary rules and with the clear goal of being accepted upstream is very interesting.

yeah, but any friction between the two will likely be Corellium's fault, the Asahi project has stated they're more than willing to accept contributors and help review any M1-related patches