Well, after beating my head against a wall for 2 days I realized this was the dumbest approach I could have taken. Instead I've passed the disks directly to the VM and everything is gravy. Quick dd benchmark of 20GB to a RAID 0 array on the thunderbolt enclosure is giving me 479 MB/s. Sorry to waste everyone's time but maybe there can be some lessons for the other stuff. Pass through the NHI devices for better success!
sudo qemu-system-x86_64 \ -vnc :0 \ -m 2048 -smp cores=2 \ -machine accel=kvm \ -drive file=/usr/share/OVMF/x64/OVMF_CODE.fd,if=pflash,format=raw,unit=0,readonly=on \ -drive file=/home/chuwi-admin/OVMF_VARS.fd,if=pflash,format=raw,unit=1 \ -drive format=raw,file=/dev/sdf \ -drive file=/dev/disk/by-path/pci-0000:03:00.0-ata-1,if=none,format=raw,aio=native,cache=none,id=hd0 \ -device virtio-blk-pci,scsi=off,drive=hd0 \ -drive file=/dev/disk/by-path/pci-0000:03:00.0-ata-2,if=none,format=raw,aio=native,cache=none,id=hd1 \ -device virtio-blk-pci,scsi=off,drive=hd1 \ -drive file=/dev/disk/by-path/pci-0000:04:00.0-ata-1,if=none,format=raw,aio=native,cache=none,id=hd2 \ -device virtio-blk-pci,scsi=off,drive=hd2 \ -drive file=/dev/disk/by-path/pci-0000:04:00.0-ata-2,if=none,format=raw,aio=native,cache=none,id=hd3 \ -device virtio-blk-pci,scsi=off,drive=hd3
_______________________________________________ vfio-users mailing list vfio-users@redhat.com https://listman.redhat.com/mailman/listinfo/vfio-users