https://intel-gfx-ci.01.org/tree/drm-tip/drmtip_394/fi-cml-s/igt@perf_pmu@cpu-hotplug.html IGT-Version: 1.24-g5eef4d167 (x86_64) (Linux: 5.4.0-rc4-g5d2161b33903-drmtip_394+ x86_64) Starting subtest: cpu-hotplug Test requirement not met in function cpu_hotplug, file ../tests/perf_pmu.c:1257: Test requirement: !(buf == 's') Subtest cpu-hotplug: SKIP (0.510s)
The CI Bug Log issue associated to this bug has been updated. ### New filters associated * CML: igt@perf_pmu@cpu-hotplug - skip - Test requirement: !(buf == 's'), SKIP - https://intel-gfx-ci.01.org/tree/drm-tip/drmtip_394/fi-cml-s/igt@perf_pmu@cpu-hotplug.html
This test attempts to toggle the offline/online status of CPUs to ensure that the hotplugging of CPUs does not impact the busyness stats returned by the perf pmu api. IGT commit ec872b7dc1 ("tests/perf_pmu: Handle CPU hotplug failures better") mentions "CPU hotplug, especially CPU0, can be flaky on commodity hardware" and that's what we're running into here. For whatever reason, the kernel is not able to take CPU0 offline, so the test prints the igt_warn() message and then skips. The behavior here is not related to graphics at all, just the inability to offline CPU0 on the system, so there should be no impact to end users. As such, marking priority/importance as low. The inability to turn off the CPU is a semi-expected behavior on some platforms, so going forward it would probably be best to downgrade the message to an igt_info() so that it won't get flagged by CI; there isn't anything we can/should do in response to this message.
-- GitLab Migration Automatic Message -- This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity. You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/drm/intel/issues/554.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.