Posts
508
Following
480
Followers
394
Linux Kernel security developer, working for Microsoft. Also W7TXT. Views are my own.
Topics: #Linux #kernel #security, #amateurradio, #RF, #hamradio, #electronics, #science, #radioastronomy, #physics, #space, #arduino.

📡 https://w7txt.net/
🐧 https://blog.namei.org/
☠️ https://www.facebook.com/w7txt


Collected various bits of 80s surplus over a couple of years and now I can “see” signals from DC to sub-millimeter (300 GHz+), in theory at least. This is a 25 GHz signal.
1
3
13
"Looks like there aren't many options for WJ-340 VLF receiver. Consider this used farm tractor or a pair of flannel pajamas."
0
0
1
DIY window feed through for running RF cables from inside to outside. You can buy these but I had the parts.
2
1
7

Good read about cancellations in Async Rust https://sunshowers.io/posts/cancelling-async-rust/

This blogpost covers a lot of the problems in this space very well. I like how it emphasizes definitions and semantics, shows examples of bugs they came across to, and list some practical recommendations.

The "double-edge sword" of cancellation in Rust is spot on: on one hand, it's cool how "easy" and simple it is to cancel future in Rust -- you *drop* it. Especially given how non-trivial it can be in synchronous environments.

However, this also means it's far too easy to *silently* drop a future. This is scary on its own, and combined with the fact that cancellation of parent futures propagates down to child futures (because of the single ownership model) makes it much more concerning.

This post also links another (much older, Feb 2024) blogpost that covers this space well, elaborating on async clean-up problems and mentions linear types as a solution https://without.boats/blog/asynchronous-clean-up/

0
1
1

Last week 2025 happened in Berlin, and it was a blast as usual, thanks to all the amazing speakers!

Talks are already available on YT: https://www.youtube.com/playlist?list=PLWYdJViL9EiqX25Rmu49FTDDnzT30wvqL

1
4
1

At the beginning of this month we wrapped up OSHWA’s first ever Open Healthware Conference. For two days we got to hear from incredible creators within the Open Healthware community, see demos of devices meant to better the medical world, and meet new folks from the community.

It was incredible! We can’t say thank you enough but we can try! From NSF, to speakers, to in person and online attendees - thank you, thank you, thank you!!

https://oshwa.org/announcements/open-healthware-conference-wrap-up/

0
1
1

All Systems Go! is almost here! Next Tuesday, Sept 30th, will kick off 2 days of talks and mingling around foundational user-space Linux technologies.
📆 Schedule is here ➡️ https://cfp.all-systems-go.io/all-systems-go-2025/schedule/
🎟️ Tickets are still available ➡️ https://ti.to/all-systems-go/all-systems-go-2025

0
9
1

David Chisnall (*Now with 50% more sarcasm!*)

Edited 10 days ago

Anyone going to be at SOSP in a couple of weeks? I will be giving a keynote at the KISV workshop on Monday morning about how CHERI changes how you think about OS design, and then we'll be presenting our paper on the CHERIoT RTOS first thing on Tuesday morning (and at the poster session).

0
1
1
Did you know you can watch @KernelRecipes videos already by seeking in the live streams which are currently online?

https://www.youtube.com/@KernelRecipes/streams
0
1
5
Running make -j32 to build a kernel on my new work workstation. Takes about a minute, in total silence.
1
1
6
@vilmibm people take it for granted now that you can just click on things, but it took decades for the idea to take hold & become ubiquitous. Most of it was utterly awful until the web came along.

RE: https://tiny.tilde.website/@vilmibm/115177383467463329
0
0
0
Edited 28 days ago
Interesting paper: "Enter, Exit, Page Fault, Leak: Testing Isolation Boundaries for Microarchitectural Leaks"

https://www.microsoft.com/en-us/research/publication/enter-exit-page-fault-leak-testing-isolation-boundaries-for-microarchitectural-leaks/

"We found 4 new cross-domain leaks, successfully detected all 6 known leaks possible in the configurations that we tested and reproduced 6 known flaws in patches."

e.g.:

"We discovered a microarchitectural effect on AMD3 that allows an attacker VM to selectively infer any bit from the memory of another VM, provided the victim VM has previously accessed (i.e., cached) this bit. By repeatedly exploiting this effect, the attacker VM could read the victim’s memory one bit at a time, and thus learn the contents of an arbitrary memory range used by the victim. "
0
3
3

Linux Security Summit 🐧

📢 🐧 The videos from LSS-EU 🇪🇺 2025 in Amsterdam 🇳🇱 are now up!

📺 Here is the playlist:

https://youtube.com/playlist?list=PLbzoR-pLrL6rSxIlgQx8OYw74Az63TpaB&si=6DEbDaY4GJMtIH1m
0
7
5
Learning KiCAD.



@kicad
2
0
8
Edited 1 month ago

4️⃣9️⃣ Here's the 49th post highlighting key new features of the upcoming v258 release of systemd.

One of the key features of systemd from day 1 on is socket activation, i.e. a mechanism where systemd binds sockets on behalf of services, watches them and only activates the services themselves later, possibly only at the moment they are actively used.

This has various benefits, for example reduces ahead of time cost of running a large number of services (which improves boot times).

1
5
1

📣 Reminder: Ticket prices go up in a few days for All Systems Go! 2025, the foundational user-space Linux technologies event in Berlin.
🎟️ Get yours today @ https://ti.to/all-systems-go/all-systems-go-2025
ℹ️ And more info @ https://all-systems-go.io/

0
5
1

Linux Security Summit 🐧

LSS-EU kicks off today in Amsterdam, with conference chair Elena Reshetova presenting opening remarks.

https://lsseu2025.sched.com/list/simple
1
3
2
Where am I? Wrong answers welcome.
1
0
0
Got addicted to Severance on a recent flight, but I need to subscribe to 🍏 tv to see the full season now. This dystopia is getting very ~meta~ blobcatchefskiss
0
0
3
Edited 1 month ago

AI agents can potentially gain extensive access to user data, and even write or execute arbitrary code.

OpenAI Codex CLI uses sandboxing to reduce the risk of buggy or malicious commands: https://github.com/openai/codex/pull/763

For now, it only blocks arbitrary file changes, but there’s room to strengthen protections further, and the ongoing rewrite in will help: https://github.com/openai/codex/pull/629

Landlock is designed for exactly this kind of use case, providing unprivileged and flexible access control.

0
3
1
Show older