Hacker News Reader

Previous discussion from 2 months ago: https://news.ycombinator.com/item?id=16419039 (43 comments)

Some interesting metrics on bpfilter, especially when combined with smart network card offloading

https://www.netronome.com/blog/bpf-ebpf-xdp-and-bpfilter-wha...

So should we now forget to disable eBPF jit in our kernel's because it is insecure (spectre attacks)? And eBPF was made even more insecure, with even more superpowers. They added tables! Not in dtrace just eBPF, because linux has so much more superpowers.

Time to switch over to a proper kernel, where such things do not happen.

BPF != eBPF

Do you have a reference expanding on that? I was just reading up on BPF/eBPF the other day (https://lwn.net/Articles/740157/) and (https://qmonnet.github.io/whirl-offload/2016/09/01/dive-into...).

From both of those I gathered that, with reference to Linux, BPF == eBPF, with the older implementation (cBPF) being deprecated.

So as the linked article is about Linux, I thought that in this case BPF == eBPF

And? Nobody is talking about BPF, only you. And lwn is confused about this, not me.

Where is the difference between BSD pf and Linux BPF implementation? eBPF aside

BPF comes to firewalls