static int
qemuCapsInitGuest(virCapsPtr caps,
virCapsPtr old_caps,
const char *hostmachine,
const struct qemu_arch_info *info,
int hvm)
{
virCapsGuestPtr guest;
int i;
int haskvm = 0;
int haskqemu = 0;
char *kvmbin = NULL;
char *binary = NULL;
time_t binary_mtime;
virCapsGuestMachinePtr *machines = NULL;
int nmachines = 0;
struct stat st;
unsigned int ncpus;
virBitmapPtr qemuCaps = NULL;
int ret = -1;
/* Check for existance of base emulator, or alternate base
* which can be used with magic cpu choice
*/
binary = virFindFileInPath(info->binary);
if (binary == NULL || !virFileIsExecutable(binary)) {
VIR_FREE(binary);
binary = virFindFileInPath(info->altbinary);
}
/* Can use acceleration for KVM/KQEMU if
* - host & guest arches match
* Or
* - hostarch is x86_64 and guest arch is i686
* The latter simply needs "-cpu qemu32"
*/
if (STREQ(info->arch, hostmachine) ||
(STREQ(hostmachine, "x86_64") && STREQ(info->arch, "i686"))) {
if (access("/dev/kvm", F_OK) == 0) {
const char *const kvmbins[] = { "/usr/libexec/qemu-kvm", /* RHEL */
"qemu-kvm", /* Fedora */
"kvm" }; /* Upstream .spec */
for (i = 0; i < ARRAY_CARDINALITY(kvmbins); ++i) {
kvmbin = virFindFileInPath(kvmbins[i]);
if (!kvmbin)
continue;
haskvm = 1;
if (!binary)
binary = kvmbin;
break;
}
}
if (access("/dev/kqemu", F_OK) == 0)
haskqemu = 1;
}
if (!binary)
return 0;
/* Ignore binary if extracting version info fails */
if (qemuCapsExtractVersionInfo(binary, info->arch, NULL, &qemuCaps) < 0) {
ret = 0;
goto cleanup;
}
if (stat(binary, &st) == 0) {
binary_mtime = st.st_mtime;
} else {
char ebuf[1024];
VIR_WARN("Failed to stat %s, most peculiar : %s",
binary, virStrerror(errno, ebuf, sizeof(ebuf)));
binary_mtime = 0;
}
if (info->machine) {
virCapsGuestMachinePtr machine;
if (VIR_ALLOC(machine) < 0) {
goto no_memory;
}
if (!(machine->name = strdup(info->machine))) {
VIR_FREE(machine);
goto no_memory;
}
nmachines = 1;
if (VIR_ALLOC_N(machines, nmachines) < 0) {
VIR_FREE(machine->name);
VIR_FREE(machine);
goto no_memory;
}
machines[0] = machine;
} else {
int probe = 1;
if (old_caps && binary_mtime)
probe = !qemuCapsGetOldMachines(hvm ? "hvm" : "xen", info->arch,
info->wordsize, binary, binary_mtime,
old_caps, &machines, &nmachines);
if (probe &&
qemuCapsProbeMachineTypes(binary, &machines, &nmachines) < 0)
goto error;
}
/* We register kvm as the base emulator too, since we can
* just give -no-kvm to disable acceleration if required */
if ((guest = virCapabilitiesAddGuest(caps,
hvm ? "hvm" : "xen",
info->arch,
info->wordsize,
binary,
NULL,
nmachines,
machines)) == NULL)
goto error;
machines = NULL;
nmachines = 0;
guest->arch.defaultInfo.emulator_mtime = binary_mtime;
if (caps->host.cpu &&
qemuCapsProbeCPUModels(binary, NULL, info->arch, &ncpus, NULL) == 0 &&
ncpus > 0 &&
!virCapabilitiesAddGuestFeature(guest, "cpuselection", 1, 0))
goto error;
if (qemuCapsGet(qemuCaps, QEMU_CAPS_BOOTINDEX) &&
!virCapabilitiesAddGuestFeature(guest, "deviceboot", 1, 0))
goto error;
if (hvm) {
if (virCapabilitiesAddGuestDomain(guest,
"qemu",
NULL,
NULL,
0,
NULL) == NULL)
goto error;
if (haskqemu &&
virCapabilitiesAddGuestDomain(guest,
"kqemu",
NULL,
NULL,
0,
NULL) == NULL)
goto error;
if (haskvm) {
virCapsGuestDomainPtr dom;
if (stat(kvmbin, &st) == 0) {
binary_mtime = st.st_mtime;
} else {
char ebuf[1024];
VIR_WARN("Failed to stat %s, most peculiar : %s",
binary, virStrerror(errno, ebuf, sizeof(ebuf)));
binary_mtime = 0;
}
if (!STREQ(binary, kvmbin)) {
int probe = 1;
if (old_caps && binary_mtime)
probe = !qemuCapsGetOldMachines("hvm", info->arch, info->wordsize,
kvmbin, binary_mtime,
old_caps, &machines, &nmachines);
if (probe &&
qemuCapsProbeMachineTypes(kvmbin, &machines, &nmachines) < 0)
goto error;
}
if ((dom = virCapabilitiesAddGuestDomain(guest,
"kvm",
kvmbin,
NULL,
nmachines,
machines)) == NULL) {
goto error;
}
machines = NULL;
nmachines = 0;
dom->info.emulator_mtime = binary_mtime;
}
} else {
if (virCapabilitiesAddGuestDomain(guest,
"kvm",
NULL,
NULL,
0,
NULL) == NULL)
goto error;
}
if (info->nflags) {
for (i = 0 ; i < info->nflags ; i++) {
if (virCapabilitiesAddGuestFeature(guest,
info->flags[i].name,
info->flags[i].default_on,
info->flags[i].toggle) == NULL)
goto error;
}
}
ret = 0;
cleanup:
if (binary == kvmbin) {
/* don't double free */
VIR_FREE(binary);
} else {
VIR_FREE(binary);
VIR_FREE(kvmbin);
}
qemuCapsFree(qemuCaps);
return ret;
no_memory:
virReportOOMError();
error:
virCapabilitiesFreeMachines(machines, nmachines);
goto cleanup;
}
int
qemuMigrationWaitForCompletion(struct qemud_driver *driver, virDomainObjPtr vm)
{
int ret = -1;
int status;
unsigned long long memProcessed;
unsigned long long memRemaining;
unsigned long long memTotal;
qemuDomainObjPrivatePtr priv = vm->privateData;
priv->jobInfo.type = VIR_DOMAIN_JOB_UNBOUNDED;
while (priv->jobInfo.type == VIR_DOMAIN_JOB_UNBOUNDED) {
/* Poll every 50ms for progress & to allow cancellation */
struct timespec ts = { .tv_sec = 0, .tv_nsec = 50 * 1000 * 1000ull };
struct timeval now;
int rc;
const char *job;
switch (priv->jobActive) {
case QEMU_JOB_MIGRATION_OUT:
job = _("migration job");
break;
case QEMU_JOB_SAVE:
job = _("domain save job");
break;
case QEMU_JOB_DUMP:
job = _("domain core dump job");
break;
default:
job = _("job");
}
if (!virDomainObjIsActive(vm)) {
qemuReportError(VIR_ERR_INTERNAL_ERROR, _("%s: %s"),
job, _("guest unexpectedly quit"));
goto cleanup;
}
if (priv->jobSignals & QEMU_JOB_SIGNAL_CANCEL) {
priv->jobSignals ^= QEMU_JOB_SIGNAL_CANCEL;
VIR_DEBUG0("Cancelling job at client request");
qemuDomainObjEnterMonitorWithDriver(driver, vm);
rc = qemuMonitorMigrateCancel(priv->mon);
qemuDomainObjExitMonitorWithDriver(driver, vm);
if (rc < 0) {
VIR_WARN0("Unable to cancel job");
}
} else if (priv->jobSignals & QEMU_JOB_SIGNAL_SUSPEND) {
priv->jobSignals ^= QEMU_JOB_SIGNAL_SUSPEND;
VIR_DEBUG0("Pausing domain for non-live migration");
if (qemuMigrationSetOffline(driver, vm) < 0)
VIR_WARN0("Unable to pause domain");
} else if (priv->jobSignals & QEMU_JOB_SIGNAL_MIGRATE_DOWNTIME) {
unsigned long long ms = priv->jobSignalsData.migrateDowntime;
priv->jobSignals ^= QEMU_JOB_SIGNAL_MIGRATE_DOWNTIME;
priv->jobSignalsData.migrateDowntime = 0;
VIR_DEBUG("Setting migration downtime to %llums", ms);
qemuDomainObjEnterMonitorWithDriver(driver, vm);
rc = qemuMonitorSetMigrationDowntime(priv->mon, ms);
qemuDomainObjExitMonitorWithDriver(driver, vm);
if (rc < 0)
VIR_WARN0("Unable to set migration downtime");
} else if (priv->jobSignals & QEMU_JOB_SIGNAL_MIGRATE_SPEED) {
unsigned long bandwidth = priv->jobSignalsData.migrateBandwidth;
priv->jobSignals ^= QEMU_JOB_SIGNAL_MIGRATE_SPEED;
priv->jobSignalsData.migrateBandwidth = 0;
VIR_DEBUG("Setting migration bandwidth to %luMbs", bandwidth);
qemuDomainObjEnterMonitorWithDriver(driver, vm);
rc = qemuMonitorSetMigrationSpeed(priv->mon, bandwidth);
qemuDomainObjExitMonitorWithDriver(driver, vm);
if (rc < 0)
VIR_WARN0("Unable to set migration speed");
}
/* Repeat check because the job signals might have caused
* guest to die
*/
if (!virDomainObjIsActive(vm)) {
qemuReportError(VIR_ERR_INTERNAL_ERROR, _("%s: %s"),
job, _("guest unexpectedly quit"));
goto cleanup;
}
qemuDomainObjEnterMonitorWithDriver(driver, vm);
rc = qemuMonitorGetMigrationStatus(priv->mon,
&status,
&memProcessed,
&memRemaining,
&memTotal);
qemuDomainObjExitMonitorWithDriver(driver, vm);
if (rc < 0) {
priv->jobInfo.type = VIR_DOMAIN_JOB_FAILED;
goto cleanup;
}
if (gettimeofday(&now, NULL) < 0) {
priv->jobInfo.type = VIR_DOMAIN_JOB_FAILED;
virReportSystemError(errno, "%s",
_("cannot get time of day"));
goto cleanup;
}
priv->jobInfo.timeElapsed = timeval_to_ms(now) - priv->jobStart;
switch (status) {
case QEMU_MONITOR_MIGRATION_STATUS_INACTIVE:
priv->jobInfo.type = VIR_DOMAIN_JOB_NONE;
qemuReportError(VIR_ERR_OPERATION_FAILED,
_("%s: %s"), job, _("is not active"));
break;
case QEMU_MONITOR_MIGRATION_STATUS_ACTIVE:
priv->jobInfo.dataTotal = memTotal;
priv->jobInfo.dataRemaining = memRemaining;
priv->jobInfo.dataProcessed = memProcessed;
priv->jobInfo.memTotal = memTotal;
priv->jobInfo.memRemaining = memRemaining;
priv->jobInfo.memProcessed = memProcessed;
break;
case QEMU_MONITOR_MIGRATION_STATUS_COMPLETED:
priv->jobInfo.type = VIR_DOMAIN_JOB_COMPLETED;
ret = 0;
break;
case QEMU_MONITOR_MIGRATION_STATUS_ERROR:
priv->jobInfo.type = VIR_DOMAIN_JOB_FAILED;
qemuReportError(VIR_ERR_OPERATION_FAILED,
_("%s: %s"), job, _("unexpectedly failed"));
break;
case QEMU_MONITOR_MIGRATION_STATUS_CANCELLED:
priv->jobInfo.type = VIR_DOMAIN_JOB_CANCELLED;
qemuReportError(VIR_ERR_OPERATION_FAILED,
_("%s: %s"), job, _("canceled by client"));
break;
}
virDomainObjUnlock(vm);
qemuDriverUnlock(driver);
nanosleep(&ts, NULL);
qemuDriverLock(driver);
virDomainObjLock(vm);
}
cleanup:
return ret;
}
/* Prepare is the first step, and it runs on the destination host.
*
* This version starts an empty VM listening on a localhost TCP port, and
* sets up the corresponding virStream to handle the incoming data.
*/
int
qemuMigrationPrepareTunnel(struct qemud_driver *driver,
virConnectPtr dconn,
virStreamPtr st,
const char *dname,
const char *dom_xml)
{
virDomainDefPtr def = NULL;
virDomainObjPtr vm = NULL;
virDomainEventPtr event = NULL;
int ret = -1;
int internalret;
int dataFD[2] = { -1, -1 };
virBitmapPtr qemuCaps = NULL;
qemuDomainObjPrivatePtr priv = NULL;
struct timeval now;
if (gettimeofday(&now, NULL) < 0) {
virReportSystemError(errno, "%s",
_("cannot get time of day"));
return -1;
}
/* Parse the domain XML. */
if (!(def = virDomainDefParseString(driver->caps, dom_xml,
VIR_DOMAIN_XML_INACTIVE)))
goto cleanup;
if (!qemuMigrationIsAllowed(def))
goto cleanup;
/* Target domain name, maybe renamed. */
if (dname) {
VIR_FREE(def->name);
def->name = strdup(dname);
if (def->name == NULL)
goto cleanup;
}
if (virDomainObjIsDuplicate(&driver->domains, def, 1) < 0)
goto cleanup;
if (!(vm = virDomainAssignDef(driver->caps,
&driver->domains,
def, true))) {
/* virDomainAssignDef already set the error */
goto cleanup;
}
def = NULL;
priv = vm->privateData;
if (qemuDomainObjBeginJobWithDriver(driver, vm) < 0)
goto cleanup;
priv->jobActive = QEMU_JOB_MIGRATION_OUT;
/* Domain starts inactive, even if the domain XML had an id field. */
vm->def->id = -1;
if (pipe(dataFD) < 0 ||
virSetCloseExec(dataFD[0]) < 0) {
virReportSystemError(errno, "%s",
_("cannot create pipe for tunnelled migration"));
goto endjob;
}
/* check that this qemu version supports the interactive exec */
if (qemuCapsExtractVersionInfo(vm->def->emulator, vm->def->os.arch,
NULL, &qemuCaps) < 0) {
qemuReportError(VIR_ERR_INTERNAL_ERROR,
_("Cannot determine QEMU argv syntax %s"),
vm->def->emulator);
goto endjob;
}
/* Start the QEMU daemon, with the same command-line arguments plus
* -incoming stdin (which qemu_command might convert to exec:cat or fd:n)
*/
internalret = qemuProcessStart(dconn, driver, vm, "stdin", true, dataFD[1],
NULL, VIR_VM_OP_MIGRATE_IN_START);
if (internalret < 0) {
qemuAuditDomainStart(vm, "migrated", false);
/* Note that we don't set an error here because qemuProcessStart
* should have already done that.
*/
if (!vm->persistent) {
virDomainRemoveInactive(&driver->domains, vm);
vm = NULL;
}
goto endjob;
}
if (virFDStreamOpen(st, dataFD[0]) < 0) {
qemuAuditDomainStart(vm, "migrated", false);
qemuProcessStop(driver, vm, 0);
if (!vm->persistent) {
if (qemuDomainObjEndJob(vm) > 0)
virDomainRemoveInactive(&driver->domains, vm);
vm = NULL;
}
virReportSystemError(errno, "%s",
_("cannot pass pipe for tunnelled migration"));
goto endjob;
}
qemuAuditDomainStart(vm, "migrated", true);
event = virDomainEventNewFromObj(vm,
VIR_DOMAIN_EVENT_STARTED,
VIR_DOMAIN_EVENT_STARTED_MIGRATED);
ret = 0;
endjob:
if (vm &&
qemuDomainObjEndJob(vm) == 0)
vm = NULL;
/* We set a fake job active which is held across
* API calls until the finish() call. This prevents
* any other APIs being invoked while incoming
* migration is taking place
*/
if (vm &&
virDomainObjIsActive(vm)) {
priv->jobActive = QEMU_JOB_MIGRATION_IN;
priv->jobInfo.type = VIR_DOMAIN_JOB_UNBOUNDED;
priv->jobStart = timeval_to_ms(now);
}
cleanup:
qemuCapsFree(qemuCaps);
virDomainDefFree(def);
VIR_FORCE_CLOSE(dataFD[0]);
VIR_FORCE_CLOSE(dataFD[1]);
if (vm)
virDomainObjUnlock(vm);
if (event)
qemuDomainEventQueue(driver, event);
qemuDriverUnlock(driver);
return ret;
}
