Re: [libvirt] [PATCH] Support reboots with the QEMU driver

For controlled shutdown we issue a 'system_powerdown' command to the QEMU monitor. This triggers an ACPI event which (most) guest OS wire up to a controlled shutdown. There is no equiv ACPI event to trigger a controlled reboot. This patch attempts to fake a reboot. - In qemuDomainObjPrivatePtr we have a bool fakeReboot flag. - The virDomainReboot method sets this flag and then triggers a normal 'system_powerdown'. - The QEMU process is started with '-no-shutdown' so that the guest CPUs pause when it powers off the guest - When we receive the 'POWEROFF' event from QEMU JSON monitor if fakeReboot is not set we invoke the qemuProcessKill command and shutdown continues normally - If fakeReboot was set, we spawn a background thread which issues 'system_reset' to perform a warm reboot of the guest hardware. Then it issues 'cont' to start the CPUs again * src/qemu/qemu_command.c: Add -no-shutdown flag if we have JSON support * src/qemu/qemu_domain.h: Add 'fakeReboot' flag to qemuDomainObjPrivate struct * src/qemu/qemu_driver.c: Fake reboot using the system_powerdown command if JSON support is available * src/qemu/qemu_monitor.c, src/qemu/qemu_monitor.h, src/qemu/qemu_monitor_json.c, src/qemu/qemu_monitor_json.h, src/qemu/qemu_monitor_text.c, src/qemu/qemu_monitor_text.h: Add binding for system_reset command * src/qemu/qemu_process.c: Reset the guest & start CPUs if fakeReboot is set --- src/qemu/qemu_command.c | 7 +++ src/qemu/qemu_domain.h | 2 + src/qemu/qemu_driver.c | 65 ++++++++++++++++++++++++++++- src/qemu/qemu_monitor.c | 19 ++++++++ src/qemu/qemu_monitor.h | 1 + src/qemu/qemu_monitor_json.c | 19 ++++++++ src/qemu/qemu_monitor_json.h | 1 + src/qemu/qemu_monitor_text.c | 13 ++++++ src/qemu/qemu_monitor_text.h | 1 + src/qemu/qemu_process.c | 95 +++++++++++++++++++++++++++++++++++++++++- 10 files changed, 220 insertions(+), 3 deletions(-) diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index 6346243..9ae1409 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -3207,6 +3207,13 @@ qemuBuildCommandLine(virConnectPtr conn, def->onReboot != VIR_DOMAIN_LIFECYCLE_RESTART) virCommandAddArg(cmd, "-no-reboot"); + /* If JSON monitor is enabled, we can receive an event + * when QEMU stops. If we use no-shutdown, then we can + * watch for this event and do a soft/warm reboot. + */ + if (monitor_json) + virCommandAddArg(cmd, "-no-shutdown"); + if (!(def->features & (1 << VIR_DOMAIN_FEATURE_ACPI))) virCommandAddArg(cmd, "-no-acpi"); diff --git a/src/qemu/qemu_domain.h b/src/qemu/qemu_domain.h index bacf5b5..b8469b6 100644 --- a/src/qemu/qemu_domain.h +++ b/src/qemu/qemu_domain.h @@ -89,6 +89,8 @@ struct _qemuDomainObjPrivate { virBitmapPtr qemuCaps; char *lockState; + + bool fakeReboot; }; struct qemuDomainWatchdogEvent diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 853c84c..4c827cf 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -1429,7 +1429,7 @@ cleanup: } -static int qemudDomainShutdown(virDomainPtr dom) { +static int qemuDomainShutdown(virDomainPtr dom) { struct qemud_driver *driver = dom->conn->privateData; virDomainObjPtr vm; int ret = -1; @@ -1461,6 +1461,8 @@ static int qemudDomainShutdown(virDomainPtr dom) { ret = qemuMonitorSystemPowerdown(priv->mon); qemuDomainObjExitMonitor(vm); + priv->fakeReboot = false; + endjob: if (qemuDomainObjEndJob(vm) == 0) vm = NULL; @@ -1472,11 +1474,66 @@ cleanup: } +static int qemuDomainReboot(virDomainPtr dom, unsigned int flags ATTRIBUTE_UNUSED) { + struct qemud_driver *driver = dom->conn->privateData; + virDomainObjPtr vm; + int ret = -1; + qemuDomainObjPrivatePtr priv; + + qemuDriverLock(driver); + vm = virDomainFindByUUID(&driver->domains, dom->uuid); + qemuDriverUnlock(driver); + + if (!vm) { + char uuidstr[VIR_UUID_STRING_BUFLEN]; + virUUIDFormat(dom->uuid, uuidstr); + qemuReportError(VIR_ERR_NO_DOMAIN, + _("no domain with matching uuid '%s'"), uuidstr); + goto cleanup; + } + priv = vm->privateData; + +#if HAVE_YAJL + if (qemuCapsGet(priv->qemuCaps, QEMU_CAPS_MONITOR_JSON)) { + if (qemuDomainObjBeginJob(vm) < 0) + goto cleanup; + + if (!virDomainObjIsActive(vm)) { + qemuReportError(VIR_ERR_OPERATION_INVALID, + "%s", _("domain is not running")); + goto endjob; + } + + qemuDomainObjEnterMonitor(vm); + ret = qemuMonitorSystemPowerdown(priv->mon); + qemuDomainObjExitMonitor(vm); + + priv->fakeReboot = true; + + endjob: + if (qemuDomainObjEndJob(vm) == 0) + vm = NULL; + } else { +#endif + qemuReportError(VIR_ERR_NO_SUPPORT, "%s", + _("Reboot is not supported without the JSON monitor")); +#if HAVE_YAJL + } +#endif + +cleanup: + if (vm) + virDomainObjUnlock(vm); + return ret; +} + + static int qemudDomainDestroy(virDomainPtr dom) { struct qemud_driver *driver = dom->conn->privateData; virDomainObjPtr vm; int ret = -1; virDomainEventPtr event = NULL; + qemuDomainObjPrivatePtr priv; qemuDriverLock(driver); vm = virDomainFindByUUID(&driver->domains, dom->uuid); @@ -1488,6 +1545,9 @@ static int qemudDomainDestroy(virDomainPtr dom) { goto cleanup; } + priv = vm->privateData; + priv->fakeReboot = false; + /* Although qemuProcessStop does this already, there may * be an outstanding job active. We want to make sure we * can kill the process even if a job is active. Killing @@ -8146,7 +8206,8 @@ static virDriver qemuDriver = { .domainLookupByName = qemudDomainLookupByName, /* 0.2.0 */ .domainSuspend = qemudDomainSuspend, /* 0.2.0 */ .domainResume = qemudDomainResume, /* 0.2.0 */ - .domainShutdown = qemudDomainShutdown, /* 0.2.0 */ + .domainShutdown = qemuDomainShutdown, /* 0.2.0 */ + .domainReboot = qemuDomainReboot, /* 0.9.3 */ .domainDestroy = qemudDomainDestroy, /* 0.2.0 */ .domainGetOSType = qemudDomainGetOSType, /* 0.2.2 */ .domainGetMaxMemory = qemudDomainGetMaxMemory, /* 0.4.2 */ diff --git a/src/qemu/qemu_monitor.c b/src/qemu/qemu_monitor.c index 1428921..1dafc4b 100644 --- a/src/qemu/qemu_monitor.c +++ b/src/qemu/qemu_monitor.c @@ -1096,6 +1096,25 @@ int qemuMonitorSystemPowerdown(qemuMonitorPtr mon) } +int qemuMonitorSystemReset(qemuMonitorPtr mon) +{ + int ret; + VIR_DEBUG("mon=%p", mon); + + if (!mon) { + qemuReportError(VIR_ERR_INVALID_ARG, "%s", + _("monitor must not be NULL")); + return -1; + } + + if (mon->json) + ret = qemuMonitorJSONSystemReset(mon); + else + ret = qemuMonitorTextSystemReset(mon); + return ret; +} + + int qemuMonitorGetCPUInfo(qemuMonitorPtr mon, int **pids) { diff --git a/src/qemu/qemu_monitor.h b/src/qemu/qemu_monitor.h index 3bb0269..ae74a3c 100644 --- a/src/qemu/qemu_monitor.h +++ b/src/qemu/qemu_monitor.h @@ -194,6 +194,7 @@ int qemuMonitorStartCPUs(qemuMonitorPtr mon, int qemuMonitorStopCPUs(qemuMonitorPtr mon); int qemuMonitorGetStatus(qemuMonitorPtr mon, bool *running); +int qemuMonitorSystemReset(qemuMonitorPtr mon); int qemuMonitorSystemPowerdown(qemuMonitorPtr mon); int qemuMonitorGetCPUInfo(qemuMonitorPtr mon, diff --git a/src/qemu/qemu_monitor_json.c b/src/qemu/qemu_monitor_json.c index 56ec65b..6cc6ea7 100644 --- a/src/qemu/qemu_monitor_json.c +++ b/src/qemu/qemu_monitor_json.c @@ -947,6 +947,25 @@ int qemuMonitorJSONSystemPowerdown(qemuMonitorPtr mon) } +int qemuMonitorJSONSystemReset(qemuMonitorPtr mon) +{ + int ret; + virJSONValuePtr cmd = qemuMonitorJSONMakeCommand("system_reset", NULL); + virJSONValuePtr reply = NULL; + if (!cmd) + return -1; + + ret = qemuMonitorJSONCommand(mon, cmd, &reply); + + if (ret == 0) + ret = qemuMonitorJSONCheckError(cmd, reply); + + virJSONValueFree(cmd); + virJSONValueFree(reply); + return ret; +} + + /* * [ { "CPU": 0, "current": true, "halted": false, "pc": 3227107138 }, * { "CPU": 1, "current": false, "halted": true, "pc": 7108165 } ] diff --git a/src/qemu/qemu_monitor_json.h b/src/qemu/qemu_monitor_json.h index 393d8fc..a72bf7c 100644 --- a/src/qemu/qemu_monitor_json.h +++ b/src/qemu/qemu_monitor_json.h @@ -49,6 +49,7 @@ int qemuMonitorJSONStopCPUs(qemuMonitorPtr mon); int qemuMonitorJSONGetStatus(qemuMonitorPtr mon, bool *running); int qemuMonitorJSONSystemPowerdown(qemuMonitorPtr mon); +int qemuMonitorJSONSystemReset(qemuMonitorPtr mon); int qemuMonitorJSONGetCPUInfo(qemuMonitorPtr mon, int **pids); diff --git a/src/qemu/qemu_monitor_text.c b/src/qemu/qemu_monitor_text.c index a16ea91..67333aa 100644 --- a/src/qemu/qemu_monitor_text.c +++ b/src/qemu/qemu_monitor_text.c @@ -417,6 +417,19 @@ int qemuMonitorTextSystemPowerdown(qemuMonitorPtr mon) { } +int qemuMonitorTextSystemReset(qemuMonitorPtr mon) { + char *info; + + if (qemuMonitorHMPCommand(mon, "system_reset", &info) < 0) { + qemuReportError(VIR_ERR_OPERATION_FAILED, + "%s", _("system reset operation failed")); + return -1; + } + VIR_FREE(info); + return 0; +} + + int qemuMonitorTextGetCPUInfo(qemuMonitorPtr mon, int **pids) { diff --git a/src/qemu/qemu_monitor_text.h b/src/qemu/qemu_monitor_text.h index 4fa5064..7536557 100644 --- a/src/qemu/qemu_monitor_text.h +++ b/src/qemu/qemu_monitor_text.h @@ -46,6 +46,7 @@ int qemuMonitorTextStopCPUs(qemuMonitorPtr mon); int qemuMonitorTextGetStatus(qemuMonitorPtr mon, bool *running); int qemuMonitorTextSystemPowerdown(qemuMonitorPtr mon); +int qemuMonitorTextSystemReset(qemuMonitorPtr mon); int qemuMonitorTextGetCPUInfo(qemuMonitorPtr mon, int **pids); diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index 6838417..694d8a9 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -348,12 +348,99 @@ qemuProcessHandleReset(qemuMonitorPtr mon ATTRIBUTE_UNUSED, } +/* + * Since we have the '-no-shutdown' flag set, the + * QEMU process will currently have guest OS shutdown + * and the CPUS stopped. To fake the reboot, we thus + * want todo a reset of the virtual hardware, followed + * by restart of the CPUs. This should result in the + * guest OS booting up again + */ +static void +qemuProcessFakeReboot(void *opaque) +{ + struct qemud_driver *driver = qemu_driver; + virDomainObjPtr vm = opaque; + qemuDomainObjPrivatePtr priv = vm->privateData; + virDomainEventPtr event = NULL; + int ret = -1; + VIR_DEBUG("vm=%p", vm); + qemuDriverLock(driver); + virDomainObjLock(vm); + if (qemuDomainObjBeginJob(vm) < 0) + goto cleanup; + + if (!virDomainObjIsActive(vm)) { + qemuReportError(VIR_ERR_INTERNAL_ERROR, "%s", + _("guest unexpectedly quit")); + goto endjob; + } + + qemuDomainObjEnterMonitorWithDriver(driver, vm); + if (qemuMonitorSystemReset(priv->mon) < 0) { + qemuDomainObjExitMonitorWithDriver(driver, vm); + goto endjob; + } + qemuDomainObjExitMonitorWithDriver(driver, vm); + + if (!virDomainObjIsActive(vm)) { + qemuReportError(VIR_ERR_INTERNAL_ERROR, "%s", + _("guest unexpectedly quit")); + goto endjob; + } + + if (qemuProcessStartCPUs(driver, vm, NULL, + VIR_DOMAIN_RUNNING_BOOTED) < 0) { + if (virGetLastError() == NULL) + qemuReportError(VIR_ERR_INTERNAL_ERROR, + "%s", _("resume operation failed")); + goto endjob; + } + event = virDomainEventNewFromObj(vm, + VIR_DOMAIN_EVENT_RESUMED, + VIR_DOMAIN_EVENT_RESUMED_UNPAUSED); + + ret = 0; + +endjob: + if (qemuDomainObjEndJob(vm) == 0) + vm = NULL; + +cleanup: + if (vm) { + if (ret == -1) + qemuProcessKill(vm); + if (virDomainObjUnref(vm) > 0) + virDomainObjUnlock(vm); + } + if (event) + qemuDomainEventQueue(driver, event); + qemuDriverUnlock(driver); +} + + static int qemuProcessHandleShutdown(qemuMonitorPtr mon ATTRIBUTE_UNUSED, virDomainObjPtr vm) { + qemuDomainObjPrivatePtr priv = vm->privateData;

+ VIR_DEBUG("vm=%p", vm); + virDomainObjLock(vm); - ((qemuDomainObjPrivatePtr) vm->privateData)->gotShutdown = true; + priv->gotShutdown = true; + if (priv->fakeReboot) { + virDomainObjRef(vm); + virThread th; + if (virThreadCreate(&th, + false, + qemuProcessFakeReboot, + vm) < 0) { + VIR_ERROR("Failed to create reboot thread, killing domain"); + qemuProcessKill(vm);

+ } + } else { + qemuProcessKill(vm); + } virDomainObjUnlock(vm); return 0; @@ -1920,6 +2007,11 @@ qemuProcessPrepareMonitorChr(struct qemud_driver *driver, } +/* + * Precondition: Both driver and vm must be locked, + * and a job must be active. This method will call + * {Enter,Exit}MonitorWithDriver + */ int qemuProcessStartCPUs(struct qemud_driver *driver, virDomainObjPtr vm, virConnectPtr conn, virDomainRunningReason reason) @@ -2182,6 +2274,7 @@ int qemuProcessStart(virConnectPtr conn, goto cleanup; vm->def->id = driver->nextvmid++; + priv->fakeReboot = false; /* Run an early hook to set-up missing devices */ if (virHookPresent(VIR_HOOK_DRIVER_QEMU)) {