On Wed, 2 Sep 2026 18:45:49 -0700 Jakub Kicinski wrote: > Completely unclear to me what the value of the tests you're adding is. > Please look judiciously at the generated code
Dropped all three selftest patches in v2. What they did: bring up two namespaces joined by veth, build a tunnel pair that disagrees in one specific way, send traffic and read the reason back from the skb:kfree_skb tracepoint. It is how I check that each path I convert reports what its commit message claims, and it is what caught six paths in an earlier revision of this series that still freed the skb with SKB_NOT_DROPPED_YET. That value is in writing the series, though, and does not by itself justify 346 lines in selftests/net. If such a test is wanted at all, I would rather come back to it separately, once the conversion has landed, and in a smaller form: - perf record -e skb:kfree_skb and perf script, the way openvswitch.sh already does it, instead of driving a trace instance by hand; - only the cases that are not evident from reading the code: a peer that reboots and restarts its sequence number, two ends configured with different options, plus a control case where nothing is dropped; - every case tied to the function the drop has to come from, so that the same reason reported by unrelated code cannot make it pass. Is that worth having, or do you consider drop reason coverage out of scope for selftests/net?
