shakedown.social is one of the many independent Mastodon servers you can use to participate in the fediverse.
A community for live music fans with roots in the jam scene. Shakedown Social is run by a team of volunteers (led by @clifff and @sethadam1) and funded by donations.

Administered by:

Server stats:

285
active users

#kernel

5 posts5 participants0 posts today

I've done a series that adds support for AF_UNIX sockets in coredumps. Userspace provides an AF_UNIX socket path via core_pattern and the kernel connects to it, shuts down the read side and writes the coredump to the socket.

This means no more super privileged usermode helper upcalls and makes for a very nice API experience. I captured coredumps simply via socat:

lore.kernel.org/20250430-work-

The receiver can use SO_PEERPIDFD to get a stable handle on the crashed process.

lore.kernel.orgMaking sure you're not a bot!

I wrote a book on Linux Memory Management, published by @nostarch - it's a comprehensive 1300 page exploration of Linux 6.0's memory management code, depth-first, diving into the code and REALLY explaining how things work.

The idea is to avoid hand waving as much as possible and literally explore what the kernel _actually_ does.

It's full of diagrams and careful explanations of logic including a ton of stuff you just can't find anywhere else.

It's currently available in its entirety in draft form via early access when you pre-order.

It's available at nostarch.com/linux-memory-mana

:)

Monday after spring break right back getting dragged into (gpu) memory management fun. I guess time to re-drop my tongue-in-cheek take on this:

memory management is trivial, right until you add reclaim

then it becomes impossible

'"[…] Suppose you compile an #eBPF program on #kernel version 5.3, but it fails to run on 5.4.

Why? Because each kernel version ships with its own kernel headers, which define structs and memory layouts. […]

So if a certain struct field […] sits at a different offset in another OS or kernel version, these helpers can still locate and read it correctly.

Under the hood, this is made possible by BPF CO-RE relocation information and BTF (BPF Type Format). […]"'

ebpfchirp.substack.com/p/why-d #Linux

eBPFChirp · Why Does My eBPF Program Work on One Kernel but Fail on Another?By Teodor J. Podobnik
I finally realized that I cannot implement one access pattern visualization that fits all. Hence next DAMON user-space tool release will contain programmable visualization feature. Users can program their access pattern visualization in Python code, as a script or interactively on the Python interpreter.

https://github.com/damonitor/damo/blob/d0eb41035db1a870d482a14087afda0196c5980b/USAGE.md#damo-report-access-programming-visualization

#linux #kernel #damon #damo
DAMON user-space tool. Contribute to damonitor/damo development by creating an account on GitHub.
GitHubdamo/USAGE.md at d0eb41035db1a870d482a14087afda0196c5980b · damonitor/damoDAMON user-space tool. Contribute to damonitor/damo development by creating an account on GitHub.

Any #NixOS users building a custom #kernel?

The #Linux kernel is crazy huge, I would like to start tweaking my system by using a custom kernel

- longterm 6.12 as base (less hassle, more stable)
- intel only (no amd, nvidia, wifi card drivers etc)
- modern-ish hardware only (no obscure old keyboard drivers etc)
- O3 optimizations (I think they generally make sense on an i7-11th gen)

Do you got good resources to start with this?

I've got a patchset that makes coredump install a pidfd for the coredumping task into the file descriptor of the process that is exec'd as a usermode helper during a pipe-based coredump. For example, systemd-coredump could set:

|/usr/lib/systemd/systemd-coredump %P %u %g %s %t %c %h %F

and %F will make the kernel install a pidfd as file descriptor number 3 for the coredumping task.

Barebones, untested, no commit messages yet:

web.git.kernel.org/pub/scm/lin

Useful?

web.git.kernel.orgMaking sure you're not a bot!
Continued thread

2/ There is now a video, too:

'"In this candid interview, #LinusTorvalds discusses #Git's unexpected journey from a #Linux #kernel management tool to the foundation of modern software development. Learn about the technical decisions, community contributions, and evolving philosophy behind the version control system that powers millions of projects."'

youtube.com/watch?v=sCr_gb8rdE