[Bug 1745895] Re: Unable to migrate vhost-net to virtio-1.0-capable kernel

[Thomas Huth]

The QEMU project is currently considering to move its bug tracking to another 
system. For this we need to know which bugs are still valid and which could be 
closed already. Thus we are setting older bugs to "Incomplete" now.
If you still think this bug report here is valid, then please switch the state 
back to "New" within the next 60 days, otherwise this report will be marked as 
"Expired". Or mark it as "Fix Released" if the problem has been solved with a 
newer version of QEMU already. Thank you and sorry for the inconvenience.

** Changed in: qemu
       Status: New => Incomplete

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1745895

Title:
  Unable to migrate vhost-net to virtio-1.0-capable kernel

Status in QEMU:
  Incomplete

Bug description:
  I am running QEMU 2.11 (from upstream source, not Red Hat package) on
  stock RHEL 6 and RHEL 7 kernels. Only the RHEL 7 kernel supports
  VIRTIO_F_VERSION_1 in its vhost-net driver.

  When migrating a guest using vhost-net from the RHEL 6 host to RHEL 7,
  the PCI config is rejected by QEMU on the target machine.

  A simple test case:

  1. On the RHEL 7 host, prepare for an incoming migration:

    rhel7# qemu-system-x86_64 -S -accel kvm -nographic -monitor stdio
  -nodefaults -netdev tap,id=net0,vhost=on,script=no,downscript=no
  -device virtio-net-pci,netdev=net0,mac=54:52:00:ff:ff:ff -incoming
  tcp:0.0.0.0:12345

  2. On the RHEL 6 host, start a guest and migrate it to the RHEL 7
  host:

    rhel6# qemu-system-x86_64 -S -accel kvm -nographic -monitor stdio 
-nodefaults -netdev tap,id=net0,vhost=on,script=no,downscript=no -device 
virtio-net-pci,netdev=net0,mac=54:52:00:ff:ff:ff
  QEMU 2.11.0 monitor - type 'help' for more information
    (qemu) migrate tcp:rhel7:12345

  The RHEL 7 QEMU errors out:

    qemu-system-x86_64: get_pci_config_device: Bad config data: i=0x20 read: 0 
device: c cmask: ff wmask: 0 w1cmask:0
    qemu-system-x86_64: Failed to load PCIDevice:config
    qemu-system-x86_64: Failed to load virtio-net:virtio
    qemu-system-x86_64: error while loading state for instance 0x0 of device 
'0000:00:02.0/virtio-net'
    qemu-system-x86_64: load of migration failed: Invalid argument

  If I start the source QEMU with vhost=off, or the target QEMU with
  disable-modern=true, the migration is successful.

  My hunch here is that the target QEMU prepares the PCI device to
  support VIRTIO_F_VERSION_1, as that's available in the kernel there,
  but then fails to (or does not know to) disable this during the
  migration.

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1745895/+subscriptions