tcp-ao/connect-deny-ipv6 is flaky

Hi Dmitry!

Looks like around Nov 14th TCP AO connect-deny-ipv6 test has gotten quite flaky. See:

https://netdev.bots.linux.dev/contest.html?test=connect-deny-ipv6

Spot checking a few looks like the failures are on test case 22:

# # 325[lib/ftrace-tcp.c:427] trace event filter tcp_ao_key_not_found [2001:db8:1::1:-1 => 2001:db8:254::1:7010, L3index 0, flags: !FS!R!P!., keyid: 100, rnext: 100, maclen: -1, sne: -1] = 0 # # 325[lib/ftrace-tcp.c:427] trace event filter tcp_hash_ao_required [2001:db8:1::1:-1 => 2001:db8:254::1:7011, L3index 0, flags: !FS!R!P!., keyid: -1, rnext: -1, maclen: -1, sne: -1] = 1 # # 325[lib/ftrace-tcp.c:427] trace event filter tcp_ao_mismatch [2001:db8:1::1:-1 => 2001:db8:254::1:7012, L3index 0, flags: !FS!R!P!., keyid: 100, rnext: 100, maclen: -1, sne: -1] = 1 # # 325[lib/ftrace-tcp.c:427] trace event filter tcp_ao_key_not_found [2001:db8:1::1:-1 => 2001:db8:254::1:7013, L3index 0, flags: !FS!R!P!., keyid: 100, rnext: 100, maclen: -1, sne: -1] = 1 # # 325[lib/ftrace-tcp.c:427] trace event filter tcp_ao_synack_no_key [2001:db8:254::1:7014 => :::0, L3index -1, flags: , keyid: 100, rnext: 100, maclen: -1, sne: -1] = 1 # # 325[lib/ftrace-tcp.c:427] trace event filter tcp_ao_wrong_maclen [2001:db8:1::1:-1 => 2001:db8:254::1:7015, L3index 0, flags: !FS!R!P!., keyid: 100, rnext: 100, maclen: -1, sne: -1] = 1 # # 325[lib/ftrace-tcp.c:427] trace event filter tcp_ao_key_not_found [2001:db8:1::1:-1 => 2001:db8:254::1:7016, L3index 0, flags: !FS!R!P!., keyid: 100, rnext: 100, maclen: -1, sne: -1] = 1

# not ok 22 Some trace events were expected, but didn't occur

Nothing looks obviously relevant in:

git log --oneline --since='Nov 14' --until='Nov 18'

I have installed virtiofsd on the test setup around Nov 9th (which speeds up tests) but that also doesn't align very well..

I'll filter that test out of our CI for now, please TAL when you have time.

Hi Jakub,

Yeah, I've noticed it as well.
Going to look into that.

Thanks,
Dmitry