Description of problem: Systemd’s tests run during the rpm build fail on baremetal Power 8 system when executed as root. Failing test is test-sd-device-monitor. Version-Release number of selected component (if applicable): systemd-251.2-1.fc37 How reproducible: Always Steps to Reproduce: 1.rpm -i systemd-251.2-1.fc37.src.rpm 2.cd ~/rpmbuild/SPECS/systemd.spec 4.dnf builddep ./systemd.spec 3.rpmbuild -ba ./systemd.spec Actual results: Build fails with 428/1065 systemd / test-sd-device-monitor FAIL 2.00s killed by signal 6 SIGABRT >>> SYSTEMD_LANGUAGE_FALLBACK_MAP=/root/rpmbuild/BUILD/systemd-stable-251.2/src/locale/language-fallback-map PATH=/root/rpmbuild/BUILD/systemd-stable-251.2/redhat-linux-build:/root/.local/bin:/root/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin MALLOC_PERTURB_=210 /root/rpmbuild/BUILD/systemd-stable-251.2/redhat-linux-build/test-sd-device-monitor ――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――― ✀ ――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――― stderr: /* test_receive_device_fail */ lo: /* test_send_receive_one(subsystem_filter=false, tag_filter=false, use_bpf=false) */ lo: /* test_send_receive_one(subsystem_filter=true, tag_filter=false, use_bpf=false) */ lo: /* test_send_receive_one(subsystem_filter=false, tag_filter=true, use_bpf=false) */ lo: /* test_send_receive_one(subsystem_filter=true, tag_filter=true, use_bpf=false) */ lo: /* test_send_receive_one(subsystem_filter=true, tag_filter=false, use_bpf=true) */ lo: /* test_send_receive_one(subsystem_filter=false, tag_filter=true, use_bpf=true) */ lo: /* test_send_receive_one(subsystem_filter=true, tag_filter=true, use_bpf=true) */ lo: /* test_subsystem_filter */ lo: Sending device subsystem:net syspath:/sys/devices/virtual/net/lo lo: /* test_tag_filter */ lo: Sending device syspath:/sys/devices/virtual/net/lo lo: /* test_sysattr_filter(ifindex) */ lo: Sending device syspath:/sys/devices/virtual/net/lo lo: /* test_sd_device_monitor_filter_remove */ lo: /* test_send_receive_one(subsystem_filter=false, tag_filter=false, use_bpf=false) */ sda: /* test_send_receive_one(subsystem_filter=false, tag_filter=false, use_bpf=false) */ sda: /* test_send_receive_one(subsystem_filter=true, tag_filter=false, use_bpf=false) */ sda: /* test_send_receive_one(subsystem_filter=false, tag_filter=true, use_bpf=false) */ sda: /* test_send_receive_one(subsystem_filter=true, tag_filter=true, use_bpf=false) */ sda: /* test_send_receive_one(subsystem_filter=true, tag_filter=false, use_bpf=true) */ sda: /* test_send_receive_one(subsystem_filter=false, tag_filter=true, use_bpf=true) */ sda: /* test_send_receive_one(subsystem_filter=true, tag_filter=true, use_bpf=true) */ sda: /* test_parent_filter */ Assertion 'device_monitor_send_device(monitor_server, monitor_client, d) >= 0' failed at src/libsystemd/sd-device/test-sd-device-monitor.c:240, function test_parent_filter(). Aborting. Expected results: All tests passing. Additional info: I have been able to reproduce this issue only on baremetal Power8 system. I have tried both with f35 and f36 based environments, both fail in the same way. To note Power9 and VMs, LPARs(even Power 8 based) were not affected by this failure. Also running unprivileged is not triggering the failure, if I have understood the test correctly it is skipped in this case.
This bug appears to have been reported against 'rawhide' during the Fedora Linux 37 development cycle. Changing version to 37.
I hope this is fixed by https://github.com/systemd/systemd/commit/8164957461ea5890aa2b6b8f397babd8840f8804. Could you test the commit?