Do you use preemptive kernel or low latency one? I have kernel 5.4.91 preemptive and haven't experienced any of this issues with RT priority qemu. I didn't run stress, but some of the games are cpu hungry and load all cpu cores. A few differences - I don't use libvirt and run qemu from command line and hugetlb for vm ram and give vm 10 cpu threads out of 16.
# Allow realtime tasks to monopolize CPU echo -1 > /proc/sys/kernel/sched_rt_runtime_us This will not help the host, but help VM responsiveness and audio. I have much better audio with 'usb-audio' qemu device. I also stopped pinning vcpus, run qemu via 'chrt -r 5 qemu-system ....' When I was pinning rt cpus I was tweaking host irq smp affinity list (not needed for isol_cpus). On Fri, Jan 29, 2021 at 5:05 AM Michael Slade <msl...@epic-code.com.au> wrote: > > This may not strictly be VFIO-related but it is latency-related. Feel > free to suggest another place to post. > > I have a VM which has 4 of the host's 8 cores, and has multiple PCI > devices passed through - GPU, audio, USB and SATA. > > If I give the vCPU threads RT priority (via libvirt XML), and then run > `stress -c 4` on the VM, the host would freeze. > > If I dedicate the 4 cores to the VM via isolcpus= and CPU pinning, then > the host is okay but there are still other issues, e.g. the guest's > audio glitches like crazy. > > Everything is fine (well mostly fine) if I refrain from setting RT > priority on the vCPU threads. > > What's happening here? Is this a scheduling bug? > > Host and guest are running kernel 5.4.91, and qemu is stable-5.0. > > > _______________________________________________ > vfio-users mailing list > vfio-users@redhat.com > https://www.redhat.com/mailman/listinfo/vfio-users > _______________________________________________ vfio-users mailing list vfio-users@redhat.com https://www.redhat.com/mailman/listinfo/vfio-users