Jakub reported that the rtnetlink test for the preferred lifetime of an address has become quite flaky. The issue started appearing around the 6.16 merge window in May, and the test fails with:
FAIL: preferred_lft addresses remaining
The flakiness might be related to power-saving behavior, as address expiration is handled by a "power-efficient" workqueue.
To address this, use slowwait to check more frequently whether the address still exists. This reduces the likelihood of the system entering a low-power state during the test, improving reliability.
Reported-by: Jakub Kicinski kuba@kernel.org Signed-off-by: Hangbin Liu liuhangbin@gmail.com --- tools/testing/selftests/net/rtnetlink.sh | 16 +++++++++++++--- 1 file changed, 13 insertions(+), 3 deletions(-)
diff --git a/tools/testing/selftests/net/rtnetlink.sh b/tools/testing/selftests/net/rtnetlink.sh index 2e8243a65b50..49141254065c 100755 --- a/tools/testing/selftests/net/rtnetlink.sh +++ b/tools/testing/selftests/net/rtnetlink.sh @@ -291,6 +291,17 @@ kci_test_route_get() end_test "PASS: route get" }
+check_addr_not_exist() +{ + dev=$1 + addr=$2 + if ip addr show dev $dev | grep -q $addr; then + return 1 + else + return 0 + fi +} + kci_test_addrlft() { for i in $(seq 10 100) ;do @@ -298,9 +309,8 @@ kci_test_addrlft() run_cmd ip addr add 10.23.11.$i/32 dev "$devdummy" preferred_lft $lft valid_lft $((lft+1)) done
- sleep 5 - run_cmd_grep_fail "10.23.11." ip addr show dev "$devdummy" - if [ $? -eq 0 ]; then + slowwait 5 check_addr_not_exist "$devdummy" "10.23.11." + if [ $? -eq 1 ]; then check_err 1 end_test "FAIL: preferred_lft addresses remaining" return
Hello:
This patch was applied to netdev/net-next.git (main) by Jakub Kicinski kuba@kernel.org:
On Tue, 15 Jul 2025 04:34:59 +0000 you wrote:
Jakub reported that the rtnetlink test for the preferred lifetime of an address has become quite flaky. The issue started appearing around the 6.16 merge window in May, and the test fails with:
FAIL: preferred_lft addresses remaining
The flakiness might be related to power-saving behavior, as address expiration is handled by a "power-efficient" workqueue.
[...]
Here is the summary with links: - [net] selftests: rtnetlink: fix addrlft test flakiness on power-saving systems https://git.kernel.org/netdev/net-next/c/3047957cc7c1
You are awesome, thank you!
On Tue, 15 Jul 2025 04:34:59 +0000 Hangbin Liu wrote:
Jakub reported that the rtnetlink test for the preferred lifetime of an address has become quite flaky. The issue started appearing around the 6.16 merge window in May, and the test fails with:
FAIL: preferred_lft addresses remaining
The flakiness might be related to power-saving behavior, as address expiration is handled by a "power-efficient" workqueue.
To address this, use slowwait to check more frequently whether the address still exists. This reduces the likelihood of the system entering a low-power state during the test, improving reliability.
It flaked again, after long time of being fine..
https://netdev-3.bots.linux.dev/vmksft-net/results/259361/15-rtnetlink-sh/st...
On Mon, Aug 18, 2025 at 06:30:12PM -0700, Jakub Kicinski wrote:
On Tue, 15 Jul 2025 04:34:59 +0000 Hangbin Liu wrote:
Jakub reported that the rtnetlink test for the preferred lifetime of an address has become quite flaky. The issue started appearing around the 6.16 merge window in May, and the test fails with:
FAIL: preferred_lft addresses remaining
The flakiness might be related to power-saving behavior, as address expiration is handled by a "power-efficient" workqueue.
To address this, use slowwait to check more frequently whether the address still exists. This reduces the likelihood of the system entering a low-power state during the test, improving reliability.
It flaked again, after long time of being fine..
https://netdev-3.bots.linux.dev/vmksft-net/results/259361/15-rtnetlink-sh/st...
Hmm, I think we need to know what's the current preferred_lft and valid_lft on failed test. Would you please add a debug line so we can get the info after failed?
slowwait 5 check_addr_not_exist "$devdummy" "10.23.11." if [ $? -eq 1 ]; then + ip addr show dev "$devdummy" check_err 1 end_test "FAIL: preferred_lft addresses remaining" return
Thanks Hangbin
On Tue, 19 Aug 2025 01:50:51 +0000 Hangbin Liu wrote:
Hmm, I think we need to know what's the current preferred_lft and valid_lft on failed test. Would you please add a debug line so we can get the info after failed?
slowwait 5 check_addr_not_exist "$devdummy" "10.23.11." if [ $? -eq 1 ]; then
ip addr show dev "$devdummy" check_err 1 end_test "FAIL: preferred_lft addresses remaining" return
Given that this is 2nd or 3rd time we're massaging this test case I think we should just add the appropriate debug information to the test itself. Please post a real patch?
linux-kselftest-mirror@lists.linaro.org