Moss: a Rust Linux-compatible kernel in 26,000 lines of code
112 points
6 days ago
| 4 comments
| github.com
| HN
hexagonal-sun
6 days ago
[-]
Hello!

For the past 8 months, or so, I've been working on a project to create a Linux-compatible kernel in nothing but Rust and assembly. I finally feel as though I have enough written that I'd like to share it with the community!

I'm currently targeting the ARM64 arch, as that's what I know best. It runs on qemu as well as various dev boards that I've got lying around (pi4, jetson nano, AMD Kria, imx8, etc). It has enough implemented to run most BusyBox commands on the console.

Major things that are missing at the moment: decent FS driver (only fat32 RO at the moment), and no networking support.

More info is on the github readme.

https://github.com/hexagonal-sun/moss

Comments & contributions welcome!

reply
Rochus
1 hour ago
[-]
Cool project, congrats. I like the idea with libkernel which makes debugging easier before going to "hardware". It's like the advantages of a microkernel achievable in a monolithic kernel, without the huge size of LKL, UML or rump kernels. Isn't Rust async/awat depending on runtime and OS features? Using it in the kernel sounds like an complex bootstrap challenge.
reply
kaoD
1 hour ago
[-]
Rust's async-await is executor-agnostic and runs entirely in userspace. It is just syntax-sugar for Futures as state machines, where "await points" are your states.

An executor (I think this is what you meant by runtime) is nothing special and doesn't need to be tied to OS features at all. You can poll and run futures in a single thread. It's just something that holds and runs futures to completion.

Not very different from an OS scheduler, except it is cooperative instead of preemptive. It's a drop in the ocean of kernel complexities.

reply
rcxdude
1 hour ago
[-]
Yeah, for example embassy-rs is an RTOS that uses rust async on tiny microcontrollers. You can hook task execution up to a main loop and interrupts pretty easily. (And RTIC is another, more radically simple version which also uses async but just runs everything in interrupt handlers and uses the interrupt priority and nesting capability of most micros to do the scheduling)
reply
Rochus
45 minutes ago
[-]
Interesting references, thanks. Moss seems to be doing the same thing as Embassy.
reply
Rochus
48 minutes ago
[-]
Ok, I see. I spent a lot of time with .Net VMs, where you cannot simply separate await from the heavy machinery that runs it. I now understand that in a kernel context, you don't need a complex runtime like Tokio. But you still need a way to wake the executor up when hardware does something (like a disk interrupt); but this indeed is not a runtime dependency.

EDIT: just found this source which explains in detail how it works: https://os.phil-opp.com/async-await/

reply
vlovich123
43 minutes ago
[-]
There’s got to be some complexity within the executor implementation though I imagine as I believe you have to suspend and resume execution of the calling thread which can be non-trivial.
reply
kaoD
23 minutes ago
[-]
You can implement an executor with threading to achieve parallelism, but it's not a fundamental characteristic of Future executors.

To reiterate: an executor is just something that runs Futures to completion, and Futures are just things that you can poll for a value.

See sibling comments for additional details.

reply
vlovich123
10 minutes ago
[-]
I’m aware; you’re not adding new information. I think you’re handwaving the difficulty of implementing work stealing in the kernel (interrupts and whatnot) + the mechanics of suspending/resuming the calling thread which isn’t as simple within the kernel as it is in userspace. eg you have to save all the register state at a minimum but it has to be integrated into the scheduler because the suspension has to pick a next task to execute and resume the register state for. On top of that you’ve got the added difficulty of doing this with work stealing (if you want good performance) and coordinating other CPUs/migrating threads between CPUs. Now you can use non interruptible sections but you really want to minimize those if you care about performance if I recall correctly.

Anyway - as I said. Implementing even a basic executor within the kernel (at least for something more involved than a single CPU machine) is more involved, especially if you care about getting good performance (and threading 100% comes up here as an OS concept so claiming it doesn’t belies a certain amount of unawareness of how kernels work internally and how they handle syscalls).

reply
kaoD
5 minutes ago
[-]
No. I am adding new information but I think you are stuck on your initial idea.

There's no work stealing. Async-await is cooperative multitasking. There is no suspending or resuming a calling thread. There is no saving register state. There is not even a thread.

I will re-reiterate: async-await is just a state machine.

reply
IshKebab
1 hour ago
[-]
Impressive work! Do you have any goals, other than learning and having fun?

Also how does it's design compare with Redox and Asterinas?

reply
bramadityaw
1 hour ago
[-]
shouldn't this be a ShowHN?
reply
marty-oehme
53 minutes ago
[-]
Very cool project! I do have to admit - looking far, far into the future - I am a bit scared of a Linux ABI-compatible kernel with an MIT license.
reply
viraptor
11 minutes ago
[-]
reply
juliangmp
5 minutes ago
[-]
I agree, I know a lot of people aren't huge fans of it but in the long run Linux being GPL2 was a huge factor in its success.
reply
stingraycharles
18 minutes ago
[-]
Why?
reply
meisel
41 minutes ago
[-]
Really neat. Do you have any specific long term goals for it? Eg, provide an OS distro (using Linux drivers?) to provide memory safety for security-critical contexts?

Also, are there any opportunities to make this kernel significantly faster than Linux’s?

reply
maxloh
1 hour ago
[-]
In what extent is this compatible with Linux?

Could I swap Ubuntu's or Android's kernel with this, while keeping those OSes bootable?

reply
tuyiown
55 minutes ago
[-]
While it's very legitimate question, the answer is between the lines in the README, and it mostly means that there is a user space binary compatibility for everything that is implemented.

It might seem obscure, but syscalls to get access to kernel requires a tight integration on compilation and linking. So this is their approach and this is where the compatibility really means something : since you can cross compile on another machine, they don't need the full toolchain right away. Just compile your code on a linux machine, and run it there. You're at the mercy of all missing kernel API implementations, but it looks like a very good strategy if you aim is to code a kernel, as you only have to focus on actual syscalls implementation without getting distracted by toolchain.

reply