dev-shm in ubuntu_stress_smoke_tests failed on L-6.2 openstack ppc64el instance

[Po-Hsu Lin]

This issue can be found since Lunar 6.2.0-23.23 in sru-20230515 and still exist in 6.2.0-27.28 with stress-ng updated to V0.16.01

It's only failing on our openstack PowerPC VM, test log:
 dev-shm STARTING
 dev-shm RETURNED 2
 dev-shm FAILED
 stress-ng: debug: [18017] invoked with './stress-ng -v -t 5 --dev-shm 4 --dev-shm-ops 3000 --ignite-cpu --syslog --verbose --verify --oomable' by user 0 'root'
 stress-ng: debug: [18017] stress-ng 0.16.01 g96a17a5951e2
 stress-ng: debug: [18017] system: Linux kt-l-l-gen-6-2-c2r4d20-u-stress-smk-test-ppc64el 6.2.0-27-generic #28-Ubuntu SMP Wed Jul 12 22:41:08 UTC 2023 ppc64le, glibc 2.37
 stress-ng: debug: [18017] RAM total: 3.8G, RAM free: 1.3G, swap free: 1023.9M
 stress-ng: debug: [18017] temporary file path: '/home/ubuntu/autotest/client/tmp/ubuntu_stress_smoke_test/src/stress-ng', filesystem type: ext2 (3567925 blocks available)
 stress-ng: debug: [18017] 2 processors online, 2 processors configured
 stress-ng: info: [18017] setting to a 5 secs run per stressor
 stress-ng: debug: [18017] cache allocate: using cache maximum level L1
 stress-ng: debug: [18017] CPU data cache: L1: 32K
 stress-ng: debug: [18017] cache allocate: shared cache buffer size: 32K
 stress-ng: info: [18017] dispatching hogs: 4 dev-shm
 stress-ng: debug: [18017] starting stressors
 stress-ng: debug: [18024] dev-shm: [18024] started (instance 0 on CPU 1)
 stress-ng: debug: [18017] 4 stressors started
 stress-ng: debug: [18027] dev-shm: [18027] started (instance 3 on CPU 0)
 stress-ng: debug: [18025] dev-shm: [18025] started (instance 1 on CPU 1)
 stress-ng: debug: [18026] dev-shm: [18026] started (instance 2 on CPU 0)
 stress-ng: debug: [18028] dev-shm: assuming killed by OOM killer, restarting again (instance 0)
 stress-ng: debug: [18028] dev-shm: assuming killed by OOM killer, restarting again (instance 0)
 stress-ng: debug: [18028] dev-shm: assuming killed by OOM killer, restarting again (instance 0)
 stress-ng: debug: [18028] dev-shm: assuming killed by OOM killer, restarting again (instance 0)
 stress-ng: debug: [18028] dev-shm: assuming killed by OOM killer, restarting again (instance 0)
 stress-ng: debug: [18028] dev-shm: assuming killed by OOM killer, restarting again (instance 0)
 stress-ng: debug: [18028] dev-shm: assuming killed by OOM killer, restarting again (instance 0)
 stress-ng: debug: [18024] dev-shm: [18024] exited (instance 0 on CPU 1)
 stress-ng: debug: [18017] dev-shm: [18024] terminated (success)
 stress-ng: debug: [18017] dev-shm: [18025] terminated on signal: 9 (Killed)
 stress-ng: debug: [18017] dev-shm: [18025] possibly killed by the OOM killer
 stress-ng: debug: [18017] dev-shm: [18025] terminated (success)
 stress-ng: debug: [18017] dev-shm: [18026] terminated on signal: 9 (Killed)
 stress-ng: debug: [18017] dev-shm: [18026] possibly killed by the OOM killer
 stress-ng: debug: [18017] dev-shm: [18026] terminated (success)
 stress-ng: debug: [18017] dev-shm: [18027] terminated on signal: 9 (Killed)
 stress-ng: debug: [18017] dev-shm: [18027] possibly killed by the OOM killer
 stress-ng: debug: [18017] dev-shm: [18027] terminated (success)
 stress-ng: metrc: [18017] stressor bogo ops real time usr time sys time bogo ops/s bogo ops/s CPU used per RSS Max
 stress-ng: metrc: [18017] (secs) (secs) (secs) (real time) (usr+sys time) instance (%) (KB)
 stress-ng: metrc: [18017] dev-shm 66 1.38 0.00 4.47 47.88 14.76 81.08 796672
 stress-ng: fail: [18017] dev_shm instance 1 corrupted bogo-ops counter, 6 vs 0
 stress-ng: fail: [18017] dev_shm instance 1 hash error in bogo-ops counter and run flag, 3195996450 vs 0
 stress-ng: fail: [18017] dev_shm instance 2 corrupted bogo-ops counter, 6 vs 0
 stress-ng: fail: [18017] dev_shm instance 2 hash error in bogo-ops counter and run flag, 3195996450 vs 0
 stress-ng: fail: [18017] dev_shm instance 3 corrupted bogo-ops counter, 5 vs 0
 info: 5 failures reached, aborting stress process
 stress-ng: fail: [18017] dev_shm instance 3 hash error in bogo-ops counter and run flag, 2551009289 vs 0
 stress-ng: fail: [18017] metrics-check: stressor metrics corrupted, data is compromised
 stress-ng: info: [18017] dev-shm:
 stress-ng: warn: [18017] 15 Spurious Interrupts
 stress-ng: info: [18017] passed: 4: dev-shm (4)
 stress-ng: info: [18017] failed: 0
 stress-ng: info: [18017] skipped: 0
 stress-ng: info: [18017] metrics untrustworthy: 0
 stress-ng: info: [18017] unsuccessful run completed in 5.52 secs

Please find attachment for call traces (the test invokes the OOM killer)

[Po-Hsu Lin (cypressyew)]

This failure can be found on Jammy with Openstack PowerPC VM as well.