Public bug reported: We've been starting to run the systemd integration tests on Github Actions, and noticed a soft lockup in of our VMs:
""" [ 245.130625] INFO: task khugepaged:38 blocked for more than 122 seconds. [ 245.131669] Not tainted 6.8.7-arch1-1 #1 [ 245.132371] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. [ 245.134651] INFO: task test-journal-ve:4563 blocked for more than 122 seconds. [ 245.136035] Not tainted 6.8.7-arch1-1 #1 [ 245.136921] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. [ 253.516332] systemd[1]: Unmounting efi.mount... [ 253.535696] systemd[1]: efi.mount: Deactivated successfully. [ 253.537261] systemd[1]: Unmounted efi.mount. [ 345.502294] systemd[1006]: Created slice background.slice. [ 345.504159] systemd[1006]: Starting systemd-tmpfiles-clean.service... [ 345.537927] systemd[1006]: Finished systemd-tmpfiles-clean.service. [ 368.017373] INFO: task khugepaged:38 blocked for more than 245 seconds. [ 368.019120] Not tainted 6.8.7-arch1-1 #1 [ 368.020219] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. [ 368.021719] INFO: task test-journal-ve:4563 blocked for more than 245 seconds. [ 368.023485] Not tainted 6.8.7-arch1-1 #1 [ 368.024593] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. [ 490.891682] INFO: task khugepaged:38 blocked for more than 368 seconds. [ 490.893456] Not tainted 6.8.7-arch1-1 #1 [ 490.894594] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. [ 490.896750] INFO: task test-journal-ve:4563 blocked for more than 368 seconds. [ 490.897930] Not tainted 6.8.7-arch1-1 #1 [ 490.898880] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. """ This is the same lockup as was already investigated in https://github.com/systemd/systemd-centos-ci/issues/720 and was traced to a missing backported commit in the *host* kernel, specifically d02c357e5bfa7dfd618b7b3015624beb71f58f1f (KVM: x86/mmu: Retry fault before acquiring mmu_lock if mapping is changing). See the linked issue for all the details. Can this commit please be backported to linux-azure to fix these soft lockups in VMs running on Github Actions? ** Affects: linux-azure (Ubuntu) Importance: Undecided Status: New -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/2062977 Title: Please backport d02c357e5bfa7dfd618b7b3015624beb71f58f1f To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux-azure/+bug/2062977/+subscriptions -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
