Some kernel selftests fail on linux-kvm

[Guilherme G. Piccoli]

This is a catch-all Launchpad report to list the kernel selftests failures in linux-kvm kernel.
The comments will be used to reference and discuss about the failures, pointing to individual issues.

[Guilherme G. Piccoli (gpiccoli)]

Currently, in both Groovy versions (5.8.0-1023.25 - was updates before security bump) and (5.8.0-1026.28 - current proposed), we see the following failures:

[...]
INFO | ========== Summary ===========
INFO | Sub test case: rtnetlink.sh failed. [0] [1]
INFO | Sub test case: xfrm_policy.sh failed. [0] [2] [3] [4]
INFO | Sub test case: fib_tests.sh failed. [0] [5]
INFO | Sub test case: fib-onlink-tests.sh failed. [0] [2] [5]
INFO | Sub test case: pmtu.sh failed. [0] [6]
INFO | Sub test case: fib_rule_tests.sh failed. [0] [5]
INFO | Sub test case: psock_snd.sh failed. [0] [7]
INFO | Sub test case: test_vxlan_under_vrf.sh failed. [0] [2]
INFO | Sub test case: so_txtime.sh failed. [0] [8]
INFO | Sub test case: fib_nexthops.sh failed. [0] [9]
INFO | Sub test case: altnames.sh failed. [0]
INFO | Sub test case: icmp_redirect.sh failed. [0] [10]
INFO | Sub test case: vrf-xfrm-tests.sh failed. [0] [2] [3] [4]
[...]
DEBUG| [stdout] # [10] Generic dynamic event - add/remove kprobe events [FAIL]
[...]
DEBUG| [stdout] not ok 1 selftests: ftrace: ftracetest # exit=1 [11]

The following LP bugs represent these failures:

[0] https://bugs.launchpad.net/ubuntu-kernel-tests/+bug/1870299
[1] https://bugs.launchpad.net/ubuntu/+source/linux-kvm/+bug/1812194

[2] https://bugs.launchpad.net/ubuntu-kernel-tests/+bug/1900645
[3] https://bugs.launchpad.net/ubuntu-kernel-tests/+bug/1909647
[4] https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1856010

[5] https://bugs.launchpad.net/ubuntu-kernel-tests/+bug/1926228
[6] https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1887661
[7] https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1892213
[8] https://bugs.launchpad.net/ubuntu-kernel-tests/+bug/1870106
[9] https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1905203
[10] https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1880645

[11] https://bugs.launchpad.net/ubuntu-kernel-tests/+bug/1923683

[Guilherme G. Piccoli (gpiccoli)]

Observed on F/KVM, cycle sru-20210510.

[Guilherme G. Piccoli (gpiccoli)]

Observed on G/KVM, cycle sru-20210510 .

[Guilherme G. Piccoli (gpiccoli)]

New observation: H/KVM (cycle 20210510) have seccomp_benchmark failing due to timeout. A manual run was successful.

02:38:23 DEBUG| [stdout] not ok 2 selftests: seccomp: seccomp_benchmark # TIMEOUT 120 seconds

Reference LP: https://bugs.launchpad.net/ubuntu/+source/linux-kvm/+bug/1929247

[Guilherme G. Piccoli (gpiccoli)]

Observed on H/KVM, cycle sru-20210510 .

[Guilherme G. Piccoli (gpiccoli)]

New observation:

13:41:43 DEBUG| [stdout] # ./ip_defrag: recv: payload_len = 8462 max_frag_len = 8: Resource temporarily unavailable
13:41:43 DEBUG| [stdout] not ok 18 selftests: net: ip_defrag.sh # exit=1

This is due to bug https://bugs.launchpad.net/ubuntu-kernel-tests/+bug/1908498 .

Noticed on G/KVM, cycle sru-20210531 .

[Guilherme G. Piccoli (gpiccoli)]

New observation, on H/KVM (cycle sru-20210531), LP https://bugs.launchpad.net/ubuntu/+source/linux-kvm/+bug/1932549 :

03:28:51 DEBUG| [stdout] # selftests: net: devlink_port_split.py
03:28:51 DEBUG| [stdout] # Error sending command: devlink -j dev show
03:28:51 DEBUG| [stdout] #
03:28:51 DEBUG| [stdout] # Failed to connect to devlink Netlink
03:28:51 DEBUG| [stdout] #
03:28:51 DEBUG| [stdout] # Traceback (most recent call last):
03:28:51 DEBUG| [stdout] # File "/home/ubuntu/autotest/client/tmp/ubuntu_kernel_selftests/src/linux/tools/testing/selftests/net/./devlink_port_split.py", line 277, in <module>
03:28:51 DEBUG| [stdout] # main()
03:28:51 DEBUG| [stdout] # File "/home/ubuntu/autotest/client/tmp/ubuntu_kernel_selftests/src/linux/tools/testing/selftests/net/./devlink_port_split.py", line 239, in main
03:28:51 DEBUG| [stdout] # assert stderr == ""
03:28:51 DEBUG| [stdout] # AssertionError
03:28:51 DEBUG| [stdout] not ok 43 selftests: net: devlink_port_split.py # exit=1

This was also noticed in previous cycles, not a regression per-se.