Update after ~5 weeks on the HWE 7.0.0-38 kernel: - srv06 (DL385p Gen8, Opteron 6386 SE): 7.0.0-38-generic, uptime 33 days, no oops, no panic, no soft lockups. On 7.0.0-31 the same host crashed after ~22 days. - srv04 (same hardware): 7.0.0-38-generic since 2026-10-13. I did a planned cold start on 2026-10-22 for an unrelated hardware repair (replaced the failed FBWC capacitor on the Smart Array P420i). Since then it has been up 25 days, about 34 days on 7.0.0-38 in total, with no oops or panic. - kdump and panic_on_oops=1 are active on both hosts. No crash dump was generated. - The other three AMD hosts (srv05, srv07, srv08) are still on mainline 6.13/6.15 and have been stable for 65+ days.
So far 7.0.0-38 looks stable for us. The fixes you pointed out (Viktor) seem to address the issue, or at least make it much rarer. I plan to move the remaining AMD hosts from the mainline builds to the HWE 7.0 kernel, one host at a time, so all of them run a supported Ubuntu kernel. I will keep monitoring and will report back if anything shows up. If a crash happens, I will attach the vmcore dmesg from /var/crash. Is there anything else you would like me to collect, or should the bug stay open until the fixes are confirmed over a longer period? Thanks a lot to both of you for the help. -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/2163642 Title: Kernel panic "Attempted to kill the idle task" on AMD Opteron multi- node NUMA under KVM (6.8.0-137) To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/2163642/+subscriptions -- ubuntu-bugs mailing list [email protected] https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
