[Impact]
livemigration of windows 2012 r2 instance with virtio balloon driver
from qemu 2.5(mitaka) to qemu 2.11(queens) is not working properly.
Especially instance keep moving e.g 2.5 -> 2.5 -> 2.11
Then It shows below msg from the 2nd mitaka node.
Migration: [ 94 %]error: internal error: qemu unexpectedly closed the monitor: 2020-09-07T07:45:11.799345Z qemu-system-x86_64: warning: Unknown firmware file in legacy mode: etc/msr_feature_control
2020-09-07T07:45:12.765618Z qemu-system-x86_64: VQ 2 size 0x80 < last_avail_idx 0x1 - used_idx 0x2
2020-09-07T07:45:12.765642Z qemu-system-x86_64: Failed to load virtio-balloon:virtio
2020-09-07T07:45:12.765648Z qemu-system-x86_64: error while loading state for instance 0x0 of device '0000:00:07.0/virtio-balloon'
2020-09-07T07:45:12.766483Z qemu-system-x86_64: load of migration failed: Operation not permitted
After patching for CVE-2016-5403, we did workaround with CVE-2015-5403-6.patch,
[Test Case]
Deploy 2 mitaka-staging machines kvm host
Deploy 1 queens-staging machines kvm host
Setting NFS server and client between them.
Deploy windows 2012r2 guest instance with virtio balloon driver on one of the mitaka host
Migrate it from mitaka to mitaka (it should be ok )
Migrate it from mitaka to queens ( it raises error )
I can reproduce this issue with baremetal or vm host
[Regressions]
As this patch is qemu related, current instance should be restarted to have this fix.
Also, this patch may cause failure of vm starting, migrating related to virtio drivers.
Especially Windows guest vm.
[Others]
Description: make sure vdev->vq[i].inuse never goes below 0
This is a work-around to fix live migrations after the patches for
CVE-2016-5403 were applied. The true root cause still needs to be
determined.
Origin: based on a patch by Len <email address hidden>
Bug-Ubuntu: https://bugs.launchpad.net/ubuntu/+source/qemu/+bug/1647389
Index: qemu-2.5+dfsg/hw/virtio/virtio.c
===================================================================
--- qemu-2.5+dfsg.orig/hw/virtio/virtio.c 2017-04-05 09:48:17.420025137 -0400
+++ qemu-2.5+dfsg/hw/virtio/virtio.c 2017-04-05 09:49:59.565337543 -0400
@@ -1510,6 +1510,7 @@
for (i = 0; i < num; i++) {
if (vdev->vq[i].vring.desc) {
uint16_t nheads;
+ int inuse_tmp;
nheads = vring_avail_idx(&vdev->vq[i]) - vdev->vq[i].last_avail_idx;
/* Check it isn't doing strange things with descriptor numbers. */
if (nheads > vdev->vq[i].vring.num) {
@@ -1527,12 +1528,15 @@
* Since max ring size < UINT16_MAX it's safe to use modulo
* UINT16_MAX + 1 subtraction.
*/
- vdev->vq[i].inuse = (uint16_t)(vdev->vq[i].last_avail_idx -
+ inuse_tmp = (int)(vdev->vq[i].last_avail_idx -
vring_used_idx(&vdev->vq[i]));
+
+ vdev->vq[i].inuse = (inuse_tmp < 0 ? 0 : inuse_tmp);
+
if (vdev->vq[i].inuse > vdev->vq[i].vring.num) {
- error_report("VQ %d size 0x%x < last_avail_idx 0x%x - "
+ error_report("VQ %d inuse %u size 0x%x < last_avail_idx 0x%x - "
"used_idx 0x%x",
- i, vdev->vq[i].vring.num,
+ i, vdev->vq[i].inuse, vdev->vq[i].vring.num,
vdev->vq[i].last_avail_idx,
vring_used_idx(&vdev->vq[i]));
return -1;
Hi, the same doesn't seem to trigger for Xenial->Bionic which should be what matches Mitaka-Queens right?
Did you find a way to reproduce it on Xenial->Bionic as well to justify the change to (non UCA) Xenial users?
The Error you reported sounded familiar, for just the error you report there was a discussion pointing to different fixes:
=> https:/
These changes only got into Ubuntu >=Bionic and would also be missing.
I wonder if these two changes would fix your issue as well (less code change, less chance for regression) or if they should be added on top of yours (more thorough fix of the area).
Thea are part of the upstream 2.6.1, so they were in since yakkety - if you use those of 2.6.1 they are quite likely to apply.