Are you using isolcpus or nohz_full? I've seen similar lockups (even the reset button not working) running a slim host with a Win8/NVIDIA VM + Fedora/IGD VM tuned to try to give the host the minimal resources possible, but with the VMs overlapping CPUs because I only have a quad-core. I built a kernel with various deadlock and watchdog support enabled an in the process also removed the isolcpus and nohz_full host boot options. So far I haven't yet been able to reproduce the hang while I'm watching for it, but I'm wondering if it had something to do with those kernel options.
_______________________________________________ vfio-users mailing list vfio-users@redhat.com https://www.redhat.com/mailman/listinfo/vfio-users
