qemu/backends/vhost-user.c
Haoqian He bc85aae420 vhost-user: return failure if backend crash when live migration
Live migration should be terminated if the vhost-user backend crashes
before the migration completes.

Specifically, since the vhost device will be stopped when VM is stopped
before the end of the live migration, in current implementation if the
backend crashes, vhost-user device set_status() won't return failure,
live migration won't perceive the disconnection between QEMU and the
backend.

When the VM is migrated to the destination, the inflight IO will be
resubmitted, and if the IO was completed out of order before, it will
cause IO error.

To fix this issue:
1. Add the return value to set_status() for VirtioDeviceClass.
  a. For the vhost-user device, return failure when the backend crashes.
  b. For other virtio devices, always return 0.
2. Return failure if vhost_dev_stop() failed for vhost-user device.

If QEMU loses connection with the vhost-user backend, virtio set_status()
can return failure to the upper layer, migration_completion() can handle
the error, terminate the live migration, and restore the VM, so that
inflight IO can be completed normally.

Signed-off-by: Haoqian He <haoqian.he@smartx.com>
Message-Id: <20250416024729.3289157-4-haoqian.he@smartx.com>
Tested-by: Lei Yang <leiyang@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2025-05-14 05:39:15 -04:00

193 lines
4.6 KiB
C

/*
* QEMU vhost-user backend
*
* Copyright (C) 2018 Red Hat Inc
*
* Authors:
* Marc-André Lureau <marcandre.lureau@redhat.com>
*
* This work is licensed under the terms of the GNU GPL, version 2 or later.
* See the COPYING file in the top-level directory.
*/
#include "qemu/osdep.h"
#include "qapi/error.h"
#include "qemu/error-report.h"
#include "qom/object_interfaces.h"
#include "system/vhost-user-backend.h"
#include "system/kvm.h"
#include "io/channel-command.h"
#include "hw/virtio/virtio-bus.h"
int
vhost_user_backend_dev_init(VhostUserBackend *b, VirtIODevice *vdev,
unsigned nvqs, Error **errp)
{
int ret;
assert(!b->vdev && vdev);
if (!vhost_user_init(&b->vhost_user, &b->chr, errp)) {
return -1;
}
b->vdev = vdev;
b->dev.nvqs = nvqs;
b->dev.vqs = g_new0(struct vhost_virtqueue, nvqs);
ret = vhost_dev_init(&b->dev, &b->vhost_user, VHOST_BACKEND_TYPE_USER, 0,
errp);
if (ret < 0) {
return -1;
}
return 0;
}
void
vhost_user_backend_start(VhostUserBackend *b)
{
BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(b->vdev)));
VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
int ret, i ;
if (b->started) {
return;
}
if (!k->set_guest_notifiers) {
error_report("binding does not support guest notifiers");
return;
}
ret = vhost_dev_enable_notifiers(&b->dev, b->vdev);
if (ret < 0) {
return;
}
ret = k->set_guest_notifiers(qbus->parent, b->dev.nvqs, true);
if (ret < 0) {
error_report("Error binding guest notifier");
goto err_host_notifiers;
}
b->dev.acked_features = b->vdev->guest_features;
ret = vhost_dev_start(&b->dev, b->vdev, true);
if (ret < 0) {
error_report("Error start vhost dev");
goto err_guest_notifiers;
}
/* guest_notifier_mask/pending not used yet, so just unmask
* everything here. virtio-pci will do the right thing by
* enabling/disabling irqfd.
*/
for (i = 0; i < b->dev.nvqs; i++) {
vhost_virtqueue_mask(&b->dev, b->vdev,
b->dev.vq_index + i, false);
}
b->started = true;
return;
err_guest_notifiers:
k->set_guest_notifiers(qbus->parent, b->dev.nvqs, false);
err_host_notifiers:
vhost_dev_disable_notifiers(&b->dev, b->vdev);
}
int
vhost_user_backend_stop(VhostUserBackend *b)
{
BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(b->vdev)));
VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
int ret;
if (!b->started) {
return 0;
}
ret = vhost_dev_stop(&b->dev, b->vdev, true);
if (k->set_guest_notifiers &&
k->set_guest_notifiers(qbus->parent, b->dev.nvqs, false) < 0) {
error_report("vhost guest notifier cleanup failed: %d", ret);
return -1;
}
vhost_dev_disable_notifiers(&b->dev, b->vdev);
b->started = false;
return ret;
}
static void set_chardev(Object *obj, const char *value, Error **errp)
{
VhostUserBackend *b = VHOST_USER_BACKEND(obj);
Chardev *chr;
if (b->completed) {
error_setg(errp, "Property 'chardev' can no longer be set");
return;
}
g_free(b->chr_name);
b->chr_name = g_strdup(value);
chr = qemu_chr_find(b->chr_name);
if (chr == NULL) {
error_set(errp, ERROR_CLASS_DEVICE_NOT_FOUND,
"Chardev '%s' not found", b->chr_name);
return;
}
if (!qemu_chr_fe_init(&b->chr, chr, errp)) {
return;
}
b->completed = true;
/* could call vhost_dev_init() so early message can be exchanged */
}
static char *get_chardev(Object *obj, Error **errp)
{
VhostUserBackend *b = VHOST_USER_BACKEND(obj);
Chardev *chr = qemu_chr_fe_get_driver(&b->chr);
if (chr && chr->label) {
return g_strdup(chr->label);
}
return NULL;
}
static void vhost_user_backend_class_init(ObjectClass *oc, const void *data)
{
object_class_property_add_str(oc, "chardev", get_chardev, set_chardev);
}
static void vhost_user_backend_finalize(Object *obj)
{
VhostUserBackend *b = VHOST_USER_BACKEND(obj);
g_free(b->dev.vqs);
g_free(b->chr_name);
vhost_user_cleanup(&b->vhost_user);
qemu_chr_fe_deinit(&b->chr, true);
}
static const TypeInfo vhost_user_backend_info = {
.name = TYPE_VHOST_USER_BACKEND,
.parent = TYPE_OBJECT,
.instance_size = sizeof(VhostUserBackend),
.class_init = vhost_user_backend_class_init,
.instance_finalize = vhost_user_backend_finalize,
};
static void register_types(void)
{
type_register_static(&vhost_user_backend_info);
}
type_init(register_types);