QEmu 1.7.0 cannot restore a 1.6.0 live snapshot made in qemu-system-x86_64

[Francois Gouget]

I have upgraded to QEmu 1.7.0 (Debian 1.7.0+dfsg-2) but now when I try to restore a live snapshot made in QEmu 1.6.0 (Debian 1.6.0+dfsg-1) I see that the VM boots from scratch instead of starting directly in the snapshot's running state.

Furthermore if the VM is already running and I try to revert to the snapshot again I get the following message:

$ virsh --connect qemu:///system snapshot-revert fgtbbuild wtb; echo $?
error: operation failed: Error -22 while loading VM state
1

I have test VMs with live snapshots corresponding to different testing configurations. So I typically revert the VMs in one of the live snapshots and run the tests. It would be pretty annoying to have to recreate all these live snapshots any time I upgrade QEmu bug it looks like I'll have to do it again.

This all sounds very much like bug 1123975 where QEmu 1.3 broke compatibility with previous versions live snapshots :-(

Here is the command being run by libvirt:

/usr/bin/qemu-system-x86_64 -name fgtbbuild -S -machine pc-1.1,accel=kvm,usb=off -m 512 -realtime mlock=off -smp 4,sockets=4,cores=1,threads=1 -uuid f510955c-17de-9907-1e33-dfe1ef7a08b6 -no-user-config -nodefaults -chardev socket,id=charmonitor,path=/var/lib/libvirt/qemu/fgtbbuild.monitor,server,nowait -mon chardev=charmonitor,id=monitor,mode=control -rtc base=utc -no-shutdown -device piix3-usb-uhci,id=usb,bus=pci.0,addr=0x1.0x2 -drive file=/mnt/storage1/qemu/fgtbbuild.qcow2,if=none,id=drive-virtio-disk0,format=qcow2,cache=writeback -device virtio-blk-pci,scsi=off,bus=pci.0,addr=0x5,drive=drive-virtio-disk0,id=virtio-disk0,bootindex=1 -drive if=none,id=drive-ide0-0-0,readonly=on,format=raw -device ide-cd,bus=ide.0,unit=0,drive=drive-ide0-0-0,id=ide0-0-0 -netdev tap,fd=25,id=hostnet0,vhost=on,vhostfd=26 -device virtio-net-pci,netdev=hostnet0,id=net0,mac=52:54:00:0a:3c:e8,bus=pci.0,addr=0x3 -chardev pty,id=charserial0 -device isa-serial,chardev=charserial0,id=serial0 -device usb-tablet,id=input0 -vnc 127.0.0.1:0 -device qxl-vga,id=video0,ram_size=67108864,vram_size=67108864,bus=pci.0,addr=0x2 -device intel-hda,id=sound0,bus=pci.0,addr=0x4 -device hda-duplex,id=sound0-codec0,bus=sound0.0,cad=0 -device virtio-balloon-pci,id=balloon0,bus=pci.0,addr=0x6 -loadvm wtb

ipxe-qemu 1.0.0+git-20120202.f6840ba-3
qemu 1.7.0+dfsg-2
qemu-keymaps 1.7.0+dfsg-2
qemu-slof 20130430+dfsg-1
qemu-system 1.7.0+dfsg-2
qemu-system-arm 1.7.0+dfsg-2
qemu-system-common 1.7.0+dfsg-2
qemu-system-mips 1.7.0+dfsg-2
qemu-system-misc 1.7.0+dfsg-2
qemu-system-ppc 1.7.0+dfsg-2
qemu-system-sparc 1.7.0+dfsg-2
qemu-system-x86 1.7.0+dfsg-2
qemu-user 1.7.0+dfsg-2
qemu-utils 1.7.0+dfsg-2
libvirt-bin 1.1.4-2
libvirt0 1.1.4-2
libvirtodbc0 6.1.6+dfsg-4

[Dr. David Alan Gilbert (dgilbert-h)]

Hi Francois,
  I've managed to reproduce this, in my log file (/var/log/libvirt/qemu/machinename.log) I see:

Unknown ramblock "0000:02.0/qxl.vram", cannot accept migration
qemu: warning: error while loading state for instance 0x0 of device 'ram'
qemu-system-x86_64: Error -22 while loading VM state

do you also see that unknown ramblock warning?

(I'm running on F20 using 1.6.0 and 1.7.0 qemu's built from source running minimal F20 guests)

Dave

[Dr. David Alan Gilbert (dgilbert-h)]

Hi Francois,
  I've done some more digging.
It looks like the problem you've hit is related to the same one that's fixed by:

http://lists.gnu.org/archive/html/qemu-devel/2013-11/msg00513.html

however that only fixes older restores ; there is a work around which is to pass to QEMU:

-global i440FX-pcihost.short_root_bus=1

on loading the snapshot; which can be done by editing the snapshot xml files - but is obviously a bit
messy.

[Francois Gouget (fgouget)]

Thanks for digging into this.
I am indeed getting the same ramblock error. So it's good that there appears to be a fix for it.
Also if I understand it correctly this particular issue only affects the 1.6.0 snapshots so given that most of my snapshots are still on 1.3.x a direct upgrade to 1.7+ will hopefully let me avoid the issue.

[Dr. David Alan Gilbert (dgilbert-h)]

Yes, my understanding of the bug is that 1.7+ should load your 1.3.x images and then snapshots taken on 1.7.x should be OK into the future.

I don't think there's currently a way of fixing those 1.6.0 snapshots; that workaround will let you load them in 1.7, but I think if you were then to take a snapshot on 1.7 with that flag, the snapshot would have the same problem.

[Thomas Huth (th-huth)]

Setting status to "Won't fix" since there is no good solution to this problem according to comment #4.