Timo Beckers

I'm a newbie, but this was all that's needed to make `GOOS=js GOARCH=wasm go build ./...` work: ```diff diff --git docs/examples/variables/variables_bpfel.go docs/examples/variables/variables_bpfel.go index f091b898..469e6f7b 100644 --- docs/examples/variables/variables_bpfel.go +++ docs/examples/variables/variables_bpfel.go @@...

@alban Thanks for the update! Closing this in favor of https://github.com/cilium/ebpf/pull/1757.

@lmb Isn't staticcheck a completely different linter with coverage for different aspects? Not sure if a bug shipped in the GH Action justifies getting rid of it.

Doing some PR housekeeping. I'm certainly open to changing this if there are clear benefits, so far the reasons are a little unsubstantiated. Thanks!

This was merged in https://github.com/cilium/ebpf/pull/1572. bpf2go work continuing in https://github.com/cilium/ebpf/pull/1543.

Regarding bpffs: there are other ways to plumb fds between processes: [`pidfd_getfd(2)`](https://man7.org/linux/man-pages/man2/pidfd_getfd.2.html) and [`SCM_RIGHTS`](https://github.com/ftrvxmtrx/fd/blob/master/fd.go). Since communication would already take place over a unix socket, the latter feels like the obvious...

@saiaunghlyanhtet Looks like none of our CI jobs currently generate coverage reports, so you'll have to enable those first. `run-tests.sh` used to do that, but that's now gone.

@saiaunghlyanhtet Why do coverage reports need to go into a separate directory? You're still not merging the output of the matrix jobs.

Can't repro this with 250k iterations, either. Let's close for now, we can always resurrect.

Another worthwhile optimization, but the wrong way to go about it. This is already somewhat of a spaghetti and this PR doesn't make it much better. `newProgramWithOptions` shouldn't make too...