Exquisite write-up and OP's simple writing has a motivating ring to it, and I'm now on the local used marketplace looking for pieces of tech like this :-)
postmarketOS in particular has a really good devices page [1] showing missing feature support at a glance, as well as guides for porting to new devices [2] and porting features from an outdated vendor-provided Linux fork to the upstream kernel [3].
[1] https://wiki.postmarketos.org/wiki/Devices [2] https://wiki.postmarketos.org/wiki/Porting_to_a_new_device [3] https://wiki.postmarketos.org/wiki/Mainlining
I would prefer a sort of dual-boot or just a simple ability to run linux in "android phones"
Like, if we were to build a linux phone somehow hacking it through a raspberry pi or the alikes, they would be so much more costly and subpar.
Android phones have whole globalization working on it and the only reason why they don't work is lack of drivers support/software side, something which can be worked on.
I think if society rewards something like PostMarketOS more/make it easier to install it, then things can be so great as well.
I know I can run a terminal inside android but till now it was only possible through qemu which had its issues...
https://www.androidauthority.com/android-linux-terminal-app-...
I am not sure but I have never really appreciated having linux run inside a vm, I'd much rather run something like waydroid in a postmarketOS phone than linux inside android in an ideal world technically speaking from strictly performance but we don't live in one and installing waydroid on postmarketos or even installing postmarketos can be a very huge issue whereas installing linux on android can be a single step with termux or userLand.
And the pmos wiki is severely lacking IMHO.
Maybe you won't find an issue as simple as fixing a button, though.
Every laptop I've used with linux has had a few non-functioning buttons and keys. I think you underestimate the widespread issue.
By the way, delving into obscure and hardware-specific kernel code, sometimes yields quite interesting generally-applicable problems. For example, @dougg3 did an (excellent!) series of articles about his work on bringing mainline kernel support to "Chumby 8", a somewhat obscure, but historically significant piece of hardware, and as a side-quest he stumbled into this:
https://www.downtowndougbrown.com/2024/04/why-is-my-cpu-usag...
...which is applicable to quite a few other machines.
It was, he submitted it, and that one character fix got him into the contributors file .. we were all highly amused, because that particular boss didn't program much, but found the bug during testing and it was, nevertheless, a huge win for him .. ;)
EDIT: it was a 2-character fix. ;)
https://lkml.iu.edu/2103.2/08109.html
(W., if you read this, I still love telling the story of how you found this bug..)
All over the desk.
I've never done either, so I'm not bragging or anything
I too went on this adventure with my laptop. Sadly I hit a wall while reverse engineering the ACPI stuff. With no logs, error messages or tools on the Windows side to intercept the ACPI events, I was at a loss but eventually gave up. Massive respect for managing it with your own laptop!!
I did manage to reverse engineer the keyboard's LEDs and drive them from user space! Studied the kernel to make a contribution but decided not to do so when I saw comments saying it is better to keep functionality in user space if at all possible.
Seeing your name in the Linux changelog must be awesome!
Years ago when I was a teenager I had a couple patches accepted for ksnake and gedit.
Certainly not as prestigious as a patch to Linux, but still the idea of code I'd written running on millions of PCs around the world felt amazing.
Contributing to the kernel is definitely on my list of things to eventually do.
0: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1019716
You know, back when I confused "hash function" with "encryption" and whatnot.
At the bottom, there is a timeline, and I noticed this entry with a LWN link:
> 2025-05-27: Sasha Levin selects my patch (and a few others) for backporting...
The new version of this tool (AUTOSEL) looks very interesting!
> AUTOSEL leverages modern large language models andI'm sure if I went to the source tree and asked people for a low-hanging-fruit task someone would be kind enough to guide me to get started, but it's still kind of overwhelming to a point where I've just avoided it.
I should probably should stop coming up with excuses and just do it, as I would like to do a lot more with filesystems and having an understanding of the kernel would probably help with that.
All it takes is just to check that the commit shows up in upstream projects such as Linux and anyone can see the code, the reviews and the authors email in the AUTHORS file which verify that this contribution / patch is indeed from the author who committed that change.
This is a very old form of social proof which saves lots time and makes Leetcode redundant. (Which can now be completely cheated with LLMs.)
Except, this tests for a wide range of tasks related to a typical SWE role: programming proficiency, reasoning through rigorous code-review, clean code, open-source and testing.
Given that Leetcode and Hackerrank tests can be cheated / beaten with LLMs, it not only has been gamed to death, but it tests for nothing that is related to the role.
> but imagine the lkml if every Bay Area wannabe company started soft-requiring this as a screen!
Then this saves time and the proof is on the interviewee to show high quality functional open source changes to high profile projects like the Linux kernel for example and the bar is set high on purpose.
Imagine if hundreds of candidates have "interview cheating" software installed, just to pass the coding interview. Both the interviewer and the interviewee lose.
;-)
Just kidding, very cool to see a blow by blow of landing a Linux patch. I felt similar excitement landing a mere emacs patch.
[0]: https://www.getzola.org/ [1]: https://github.com/micahkepe/radion
I also got my first and so far only kernel patch submitted years ago:
On my MacBook Pro 2010 with GeForce 320M running any OpenGL application, even glxgears, would crash because it ran out of memory. So I dug into nouveau drm code and found out that some memory related function was using wrong code path for that GPU.
It took some time to figure out how to submit a patch but it felt nice after I got it accepted.
I didn't even know about those patch sanity-checking scripts back then, they look useful for potential future patches!
Broken link points here and shows no commits: https://github.com/torvalds/linux/commits?author=vkoskiv&sin...
If I correct the date range to include the maintainer commit date (May 15th), the commit shows up: https://github.com/torvalds/linux/commits?author=vkoskiv&sin...
Also strange that if I only filter on author[1], nothing shows up. Though I suspect that dataset might only be updated periodically.
[0]: https://docs.github.com/en/account-and-profile/reference/pro...
[1]: https://github.com/torvalds/linux/commits?author=vkoskiv
Did you try to use any AI tools during their process?