On Wed, May 15, 2024 at 03:24:10PM +0200, Florian Westphal wrote: > Phil Sutter <phil@xxxxxx> wrote: > > WDYT, is something still missing I could add to the test? Also, I'm not > > sure whether I should add it to netfilter selftests as it doesn't have a > > defined failure outcome. > > Isn't the expected outcome "did not crash"? > > You could just append a test for /proc/sys/kernel/tainted, > i.e. script ran and no splat was triggered. Fair point, thanks! > As the selftests are run in regular intervals on the netdev > CI the only critical factor is total test run time, but so far > the netfilter tests are not too bad and for the much-slower-debug kernel > the script can detect this and spent fewer cycles. My script is bound by the configured iperf3 test time, so in theory it should take the same time irrespective of system performance. Cheers, Phil
