[vfio-users] vfio corruption?

[Ivan Volosyuk]

Kernel version: 4.19.86-gentoo
Qemu: 4.0.0
Command line:
+ /home/ivan/opt/bin/qemu-system-x86_64 -enable-kvm -m 8192 -mem-path
/dev/qemu-mem -cpu host,kvm=off,hv_r
elaxed,hv_spinlocks=0x1fff,hv_vapic,hv_time,hv-vendor-id=servo,hv-vpindex,hv-reset,hv-runtime,hv-crash,-hy
pervisor -smp 4,sockets=1,cores=4,threads=1 -vcpu 0,affinity=4 -vcpu
1,affinity=5 -vcpu 2,affinity=6 -vcpu
3,affinity=7 -device virtio-net-pci,netdev=eth0,mac=DE:AD:BE:EF:CE:B9
-netdev bridge,br=brkvm,id=eth0 -no
graphic -monitor telnet:127.0.0.1:1234,server,nowait -usb -device
usb-host,vendorid=0x0461,productid=0x4d1
5,bus=usb-bus.0 -usb -device
usb-host,vendorid=0x0b38,productid=0x0010,bus=usb-bus.0 -vga none
-device vfi
o-pci,host=01:00.0,addr=6.0,multifunction=on -device
vfio-pci,host=01:00.1,addr=6.1 -device vfio-pci,host=
01:00.2,addr=6.2 -soundhw ac97 -rtc base=utc,driftfix=slew -no-hpet
-global kvm-pit.lost_tick_policy=disca
rd -device ahci,id=ahci -device virtio-scsi-pci,id=scsi0 -drive
file=/ssd/win81/win81f.raw,id=disk,if=none
,format=raw,cache=none,aio=threads,discard=unmap -device
scsi-hd,bus=scsi0.0,drive=disk -drive file=/archi
ve/system/qemu/games/games.raw,format=raw,id=games_disk,if=none,cache=none,aio=threads,discard=unmap
-devi
ce scsi-hd,drive=games_disk -drive
file=/ssd/games_300/games_300.raw,format=raw,id=games_ssd_zfs_disk,if=n
one,cache=none,aio=threads,discard=unmap -device
scsi-hd,drive=games_ssd_zfs_disk -drive if=pflash,format=
raw,readonly,file=/home/ivan/qemu/edk2-ovmf-git-20190219/OVMF_CODE.fd,format=raw
-drive if=pflash,format=r
aw,file=/home/ivan/qemu/win81.vars,format=raw
++ pidof pulseaudio
+ for pid in $(pidof pulseaudio)
+ gdb --pid 7740 --batch -x /home/vm-images/scripts/mlockall
[New LWP 7827]
[New LWP 7835]
[New LWP 7836]
[New LWP 7837]

warning: .dynamic section for "/lib64/ld-linux-x86-64.so.2" is not at
the expected address (wrong library
or version mismatch?)
qemu-system-x86_64: -device
vfio-pci,host=01:00.0,addr=6.0,multifunction=on: VFIO_MAP_DMA: -14
qemu-system-x86_64: -device
vfio-pci,host=01:00.0,addr=6.0,multifunction=on:
vfio_dma_map(0x556d1f02b3d0,
0x100000, 0xbff00000, 0x7f75fc0ff000) = -14 (Bad address)
qemu-system-x86_64: -device
vfio-pci,host=01:00.0,addr=6.0,multifunction=on: VFIO_MAP_DMA: -14
qemu-system-x86_64: -device
vfio-pci,host=01:00.0,addr=6.0,multifunction=on:
vfio_dma_map(0x556d1f02b3d0,
0x100000000, 0x140000000, 0x7f76bbfff000) = -14 (Bad address)
qemu-system-x86_64: -device
vfio-pci,host=01:00.0,addr=6.0,multifunction=on: vfio 0000:01:00.0:
failed to
setup container for group 1: memory listener initialization failed for
container: Bad address

I wonder what's the errors mean? Is this some sort of kernel memory
corruption, which causes this?