Jakub Kicinski kuba@kernel.org wrote:
On Tue, 16 Jan 2024 14:21:51 -0500 Benjamin Poirier wrote:
real 13m35.065s user 0m1.657s sys 0m27.918s
The test is not cpu bound; as Jay pointed out, it spends most of its time sleeping.
Ugh, so it does multiple iterations of 118 sec?
Could you send a patch to bump the timeout to 900 or 1200 in this case?
We could also lower the interval or number of notifications; right now "peer_notif_delay 1000" puts 1 second between the ARPs in the num_grat_arp() test. I'm not sure why that value was chosen, but the peer_notify_delay is rounded to units of the miimon interval, and in this test miimon is 100 msec.
I haven't tested this at all, but something like
diff --git a/tools/testing/selftests/drivers/net/bonding/bond_options.sh b/tools/testing/selftests/drivers/net/bonding/bond_options.sh index c54d1697f439..95eb77aebc3c 100755 --- a/tools/testing/selftests/drivers/net/bonding/bond_options.sh +++ b/tools/testing/selftests/drivers/net/bonding/bond_options.sh @@ -277,7 +277,7 @@ garp_test() ip -n ${s_ns} link set ${active_slave} down
exp_num=$(echo "${param}" | cut -f6 -d ' ') - sleep $((exp_num + 2)) + sleep $((exp_num / 5 + 2))
active_slave=$(cmd_jq "ip -n ${s_ns} -d -j link show bond0" ".[].linkinfo.info_data.active_slave")
@@ -297,7 +297,7 @@ num_grat_arp() { local val for val in 10 20 30 50; do - garp_test "mode active-backup miimon 100 num_grat_arp $val peer_notify_delay 1000" + garp_test "mode active-backup miimon 100 num_grat_arp $val peer_notify_delay 200" log_test "num_grat_arp" "active-backup miimon num_grat_arp $val" done }
could substantially reduce the time to run the test. It's kind of icky with magic numbers, but that could be cleaned up.
-J
--- -Jay Vosburgh, jay.vosburgh@canonical.com