mastodontech.de ist einer von vielen unabhängigen Mastodon-Servern, mit dem du dich im Fediverse beteiligen kannst.
Offen für alle (über 16) und bereitgestellt von Markus'Blog

Serverstatistik:

1,5 Tsd.
aktive Profile

#bpf

0 Beiträge0 Beteiligte0 Beiträge heute
Improving network filtering performance with Bpfilter
(free link) https://lwn.net/SubscriberLink/1017705/ca0af831379ca794/

From servers in a data center to desktop computers, many devices communicating on a network will eventually have to filter traffic, whether for security or performance reasons. Bpfilter is a project that allows packet filtering to be done with BPF, which can be faster than other mechanisms. Even small performance improvements in this area can lead to significant gains.

#linux #kernel #bpfilter #bpf
LWN.netFaster firewalls with bpfilter From servers in a data center to desktop computers, many devices communicating on a network wi [...]

RFC 9669: BPF Instruction Set Architecture (ISA)

On a souvent envie de faire tourner des programmes à soi dans le noyau du système d'exploitation, par exemple à des fins de déboguage ou d'observation du système. Cela soulève plein de problèmes et la technique #eBPF permet, depuis de nombreuses années, de le faire avec moins de risques. Ce #RFC spécifie le jeu d'instructions eBPF.

bortzmeyer.org/9669.html

www.bortzmeyer.orgBlog Stéphane Bortzmeyer: RFC 9669: BPF Instruction Set Architecture (ISA)

|| this close to generating ebpf from Go. Which proves it can work. Still unsure how to continue and if code generation is the way to go.

As for ebpf, it's strange because there is also a huge userland component that can interact with the bpf blob ... Should that all be generated or should I aim for a function that "compiles" go code that can be interacted with in the larger program.... ? #go #bpf

Pushed a new update to github.com/miekg/ebpf/ #golang #bpf

Finally writing some code. I -still- need to write a compiler, but looks very much doable. First will use mugo and then transform that code using Go/ast and friends.

Feeling I get now is that I should generate the bpf assembler + the surrounding Go program as that only needs to serve the specifics for this bpf snippet.

Also still not sure if I want to go the dtrace route, although at this point that would mean a different input DSL

GitHubGitHub - miekg/ebpf: ebpf compiler in Go; Write Go, get ebpfebpf compiler in Go; Write Go, get ebpf. Contribute to miekg/ebpf development by creating an account on GitHub.

The mainline-wo-mergew copr of the #kernel vanilla repositories[1] for #Fedora #Linux now ship 6.12 pre-releases that have the #BPF extensible scheduler class aka #sched_ext"[2] enabled, just like Fedora Rawhide will do with the next push of a new #LinuxKernel.

Many thx to Jose Fernandez for realizing this![3]

[1] fedoraproject.org/wiki/Kernel_
[2] lwn.net/Articles/972710/
[3] gitlab.com/cki-project/kernel-

"'We show that eBPF is Turing complete, which means it can be used for any computable problem'

[…]

Implementing Game of Life entirely in #eBPF means that it is Turing complete, so any computable problem that can be solved in any programming language, can also be solved in eBPF. Those complex problems that people thought were beyond the capabilities of eBPF, like parsing application-level protocols, or terminating TLS connections, are now shown to be possible. […]"

isovalent.com/blog/post/ebpf-f #BPF

isovalent.comeBPF for Anything! - IsovalentWe show that eBPF is Turing complete, which means it can be used for any computable problem, via the game of life demo

After Linus statement ~10 days ago[1] it looked like #BPF extensible scheduler class aka the #sched_ext (SCX) patchset would be merged for #Linux 6.11. Now there are discussions to change a few things before merging it; as of now it's unclear if that will happen and/or delay things: lore.kernel.org/all/CAHk-=wg3R

Side note: Tejun a few days ago posted v7 of the patchset and created a korg tree to coordinate development: lore.kernel.org/all/2024061821

[1] fosstodon.org/@kernellogger/11

Linus commented on the BPF extensible scheduler class aka the sched_ext (SCX) patchset:

'"I honestly see no reason to delay this any more. […] my
current plan is to merge this for [#Linux #kernel] 6.11.

[…]

I'm also not a believer in the argument that has been used (multiple times) that the #BPF #scheduler would keep people from participating in scheduler development. […]"'

lore.kernel.org/all/CAHk-%3Dwg