Feel emotional, last day at Oracle done.

It was my dream job, literally being paid to do my hobby full time.

I had incredible freedom to do whatever I wanted upstream for nearly 2 years straight.

I worked with some intensely smart people like Liam Howlett and Matthew Wilcox (two memory management legends) among many other great people.

I went from a hobbyist to a core kernel maintainer with many varied contributions throughout mm.

It was a great team to be part of and I'm very grateful to have had the opportunity.

With that said, I'm very excited about what comes next!

I will remain focused on upstream and my core mm maintainership continues :)

More on that shortly!

When I first pushed back on randomly putting LLM configs in the kernel, I was excluded from the conversation.

Then when further discussion was had about it, I was excluded from the conversation.

Then when I submitted a proposal to the maintainer's summit about AI, I was rejected and excluded from the conversation there.

And when documentation was submitted to the kernel, my feedback was ignored and I was excluded from that conversation too.

There was literal press about Linus calling me an idiot for it.

When I pushed back on a person lying about using LLMs to generate code, I was attacked and contradicted and forced to concede the discussion.

And recently, in yet another discussion about LLMs in the kernel, I was excluded from that too.

At each point I've had to fight to have my point of view heard.

It feels like many people in the kernel community just want to stick their head in the sand about AI slop.

But it's not magically going away. Ignore me all you want.

So, seemingly, somebody's Fedora and proprietary-forge credentials were compromised and used by some sort of LLM-driven bot to take over a lot of Fedora bugs:

https://lwn.net/ml/all/bf38c0fd4537c2908a84b4a4b1fcec8083925918.camel@fedoraproject.org

This person is now claiming to have regained access to the accounts, but it seems that not everybody is buying it.

What a world we have made for ourselves...

Finally found a comic about myself!

Everyone wants to be tall until they take a flight

Kernel memory management microconf at LPC 2026 is announced with the opening of CFP: https://lore.kernel.org/e3dc73df-0032-4e71-ab80-21b29c6efb46@kernel.org

#linuxplumbers

Me: as suggested by willy, adopt an unusual term in order to avoid an established term with weapons connotation.

Youtube shorts algo: haha, look at this, you fool.

(~40 seconds in) - https://www.youtube.com/shorts/UkNnZgBnyqQ

After Gitlab's recent announcement I am strongly considering migrating Redox OS to Forgejo, a truly open source community maintained project.

https://about.gitlab.com/blog/gitlab-act-2/

@lwn DAMON LSF/MM/BPF discussions covered by LWN! It was a productive and fantastic time! Thanks again to the conference organizers, and huge thanks to LWN for covering the discussions!

#linux #kernel #damon

RE: https://fedi.lwn.net/@lwn/116539161262678640

Bad things can happen if you leak pointers out of RCU read-side critical sections. What can you do about it?
https://people.kernel.org/paulmck/stupid-rcu-tricks-detecting-pointer-leaks

Typical ML argument: "If I can read something legally, why can't I train an LLM on it?"

Humans are capable of reading things and later writing a similar thing that is still a copyright violation. If I go and write a book that follows the plot line of Star Wars, that's still a copyright violation, even if no text is literally the same. If I play the melody to a song on my piano and release it without the appropriate mechanical cover license, that's also a copyright violation.

The reason this does not happen often is that, as humans, we are aware that that's plagiarism and there are rules. Sometimes it happens by accident, and people still get sued and lose.

LLMs have no such awareness and routinely output things which are blatant copyright violations when appropriately prompted. That means the model weights encode that work, and therefore, are themselves a derivative work.

Your brain encodes a massive amount of copyrighted information. You are not a walking copyright violation because humans aren't data, can't be copied and distributed en masse, have human rights, etc. This is why "mind reading machines" are a classic dystopian plot point (monetizing your thoughts etc).

An LLM is not a human, does not have human rights, nor human privileges. It is data, and if it encodes copyrighted information, that's a derivative work. If you aren't following the license of the training data, that's a copyright violation.

it is quite confusing that the floor index start from zero here in Zagreb

I need a break, but I can't take off so just zoning out for 30 minutes

LSF/MM/BPF preliminary schedule [1] is shared [2]. Seems there will be two DAMON topic sessions. Time to really focus on the preparation.

[2] https://docs.google.com/spreadsheets/d/1mGEdDrWskp7Ua91jGXzquQGinorcD58DAVXhOiRp2Gg/edit?gid=1852749899#gid=1852749899
[2] https://lore.kernel.org/all/20260423-auseinander-steigen-b80ee53079a8@brauner/

#linux #kernel #damon #lsfmmbpf

when I open Mastodon, and think "hmm, there's a name and profile picture that I don't recognize. who is it?"

and then I realize: oh wait, it's @lkundrak !

the last few names I remember is "hammer smashed filesystem" and "pope of nope"

I don't find code commentary helpful unless it explains the design first, because these days the implementation is way too complicated for readers to grasp the design by reading the implementation.

Oftentimes such code commentary mechanically describes list of statements, rather than explaining the idea behind it.

A well-written document should try to describe the idea behind the implementation, rather than the implementation itself. (Yeah, that's challenging)

They don't save much time compared to directly reading the code.

(some random rant of the day)

I used think that using en/em dashes in writing is pretty elegant... until LLMs started ruining them, and now using them makes it look like it was generated by an LLM.

Someone asked me to sign their gpg key so I did. Am I doing this gpg thing right?

2006: talk is cheap, show me the code!
2026: code is cheap, show me the talk!

Red Hat unions against layoffs !

https://solidairesinformatique.org/2026/04/16/red-hat-unions-against-lay-offs/

#RedHat