Thanks for your PR.

Basically I have no question in this change, just want to add a point:

I'm not sure how this change can fix the non-interrupting issue. If ctrl-c doesn't stop the process, there must be somewhere blocking for a long term so that process has no change to reach the "cancel point" to check <-ctx.Done(). As far as I can see, this can only happen on the bpf syscall LookupAndDelete or fs IO skbw.Write and pcapw.WritePacket, but none of these could be fixed by this PR, could it?

If ctrl-c doesn't stop the process, there must be somewhere blocking for a long term so that process has no change to reach the "cancel point" to check <-ctx.Done()

In most cases, line 111 will always be executed and there is no way to break the first for loop. for example:

skbdump -i eth0

I can not terminate this process through ctrl-c.

Understood, you pointed out that the process can't reach ctx check point when two bpf syscalls keep succeeding fetching events.

You're correct.