canonical-ubuntu-qa team mailing list archive

[Jürg Häfliger <1895258@xxxxxxxxxxxxxxxxxx>]

** Tags added: kernel-daily-bug

-- 
You received this bug notification because you are a member of Canonical
Platform QA Team, which is subscribed to ubuntu-kernel-tests.
https://bugs.launchpad.net/bugs/1895258

Title:
  rtnetlink.sh in net from ubuntu_kernel_selftests is returning 1 for a
  skipped test

Status in ubuntu-kernel-tests:
  Fix Released
Status in linux package in Ubuntu:
  Fix Released
Status in linux source package in Bionic:
  Fix Released

Bug description:
  == SRU Justification ==
  The rtnetlink.sh in kselftest/net from 4.15 kernel will fail with:
   selftests: rtnetlink.sh
   ========================================
   PASS: policy routing
   PASS: route get
   PASS: preferred_lft addresses have expired
   PASS: tc htb hierarchy
   PASS: gre tunnel endpoint
   PASS: bridge setup
   PASS: ipv6 addrlabel
   PASS: set ifalias e1dfde61-e5a7-415f-8bd9-924e9796038f for test-dummy0
   PASS: vrf
   PASS: vxlan
   SKIP: fou: iproute2 too old
   SKIP: macsec: iproute2 too old
   not ok 1..10 selftests: rtnetlink.sh [FAIL]

  The return value for this script is 1.

  Look into the code, it's the kci_test_encap_fou() test that causes
  this issue:

        ip fou help 2>&1 |grep -q 'Usage: ip fou'
        if [ $? -ne 0 ];then
              echo "SKIP: fou: iproute2 too old"
              return 1
        fi

  It should at least return ksft_skip=4 (or 0 with older release that
  does not support this skip code in kselftest framework)

  == Fix ==
  * 57aefc7c226d ("selftests: net: return Kselftest Skip code for
  skipped tests")

  This patch needs to be backported into Bionic, as some newer tests
  does not exist in Bionic.

  == Test ==
  Tested directly on the affected kernel, the test will be marked as:
  not ok 1..10 selftests:  rtnetlink.sh [SKIP]

  == Regression Potential ==
  Low, changes limited to test scripts. And the kselftest framework in
  Bionic can handle this ksft_skip=4 return code properly, tests won't
  be marked as failed because of this.

  Although IMO in this case the test should be marked as PASS instead of
  SKIP in the end, as not all tests were skipped, but that's something
  we could fix with upstream in the future.

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu-kernel-tests/+bug/1895258/+subscriptions