Timo Beckers

The reason behind these checks is simple: to validate assumptions and avoid silent breakage. Without it, we'd now emit TCP Flows without flags, which still doesn't sound valid, and may...

Okay, looks like destroy events do not/no longer carry TCP flags, I'll get this fixed. I'd need to validate this, but my assumption is tcp_flags_orig/reply were included but zero in...

Ping, looks like this is still relevant.

I've been hacking on this for a few days as well, this might take a bit longer to get right. The tests are passing, but there seems to be a...

> I did not abandon this work, I was just waiting for some feedback. Sorry Francis, that's some bad communication on my part. You've addressed all my feedback nicely and...

> How do you plan to add the patches? By default, maintainers of a repo are allowed to push to PR branches in forks, unless explicitly disabled. The sidebar of...

@eiffel-fl Sorry, on PTO this week. Just gave this another try (this is after `gh pr checkout eiffel-fl:francis/ebpf-kconfig` and reworking existing commits and prepending one commit to the existing two:...

@eiffel-fl Could you try re-forking from `cilium/ebpf` instead, and recreating this PR using my branch at https://github.com/ti-mo/ebpf/tree/tb/kconfig? I've added some code that fixes the alignment of the .kconfig Datasec and...

Converted to draft as it looks like some discussion will be needed. I haven't followed the previous discussion in detail, but wouldn't it be better to agree on an approach...

@lmb Correct, we're not forcing anything downstream. Individual modules always build against the exact version as specified in `go.sum`, even when imported into larger projects. As far as I understand,...