- Devel - libvirt List Archives

[libvirt] [PATCH] daemon: Plug memory leaks
by Alex Jia 12 Apr '12

12 Apr '12

Detected by valgrind. Leaks are introduced in commit 6e6e9be. * daemon/libvirtd-config.c (macro GET_CONF_STR): fix memory leaks. How to reproduce? % make && make -C tests check TESTS=libvirtdconftest % cd tests && valgrind -v --leak-check=full ./libvirtdconftest actual result: ==11008== 185 bytes in 5 blocks are definitely lost in loss record 3 of 5 ==11008== at 0x4A05FDE: malloc (vg_replace_malloc.c:236) ==11008== by 0x39CF07F6E1: strdup (strdup.c:43) ==11008== by 0x406626: daemonConfigLoadOptions (libvirtd-config.c:438) ==11008== by 0x406800: daemonConfigLoadData (libvirtd-config.c:492) ==11008== by 0x403CCF: testCorrupt (libvirtdconftest.c:110) ==11008== by 0x404FAD: virtTestRun (testutils.c:145) ==11008== by 0x403A34: mymain (libvirtdconftest.c:219) ==11008== by 0x404687: virtTestMain (testutils.c:700) ==11008== by 0x39CF01ECDC: (below main) (libc-start.c:226) ==11008== ==11008== LEAK SUMMARY: ==11008== definitely lost: 185 bytes in 5 blocks Signed-off-by: Alex Jia <ajia(a)redhat.com> --- daemon/libvirtd-config.c | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/daemon/libvirtd-config.c b/daemon/libvirtd-config.c index 4d041f0..0a4323a 100644 --- a/daemon/libvirtd-config.c +++ b/daemon/libvirtd-config.c @@ -152,6 +152,7 @@ checkType (virConfValuePtr p, const char *filename, virReportOOMError(); \ goto error; \ } \ + VIR_FREE(data->var_name); \ } \ } while (0) -- 1.7.1

2 2

[libvirt] [test-API PATCH 0/5] Destroy class definition in utils.py
by Osier Yang 12 Apr '12

12 Apr '12

These five patches are to change utils.py into a collection of functions. And cleanup all of existing testcase to use the functions in it directly. (1) Destroy Utils class definition (2) Remove "util = utils.Utils()" (3) Substitue "util." with "utils." (4) Remove useless utils import "from utils import utils" in testcases where none of functions in utils.py is called In testcase from utils import utils ... utils.functions(...) ...

2 10

[libvirt] [PATCH 0/5] fix block job bugs relative to upstream qemu
by Eric Blake 12 Apr '12

12 Apr '12

The following 5 patches are ready for review and application to libvirt now, since they correspond to features that are currently in the qemu 1.1 tree. (I've split these less-controversial parts from v4 of my larger storage migration series [1]; so that this series has no dependency on the yet-to-be-determined 'drive-mirror' semantics). [1]https://www.redhat.com/archives/libvir-list/2012-April/msg00330.html Patch 1 is new, and matches a qemu patch taken just today. Patch 2 matches 2/18 in the block migration series, with some tweaks to defer any qemu changes till later, as well as wording improvements in the documentation to match the behavior implemented by this series. Patch 3 matches an independent patch I sent earlier today, updated to resolve review comments from that version. Patch 4 matches 3/18 in the block migration series, with modifications to now account for both RHEL 6.2 and qemu 1.1 behaviors. I've altered this so much from Adam's original that I claimed authorship, although I've left his Signed-off-by for the initial implementation. Patch 5 matches 4/18 in the block migration series, with minor tweaks due to rebasing. Later, I will repost a v5 of my storage migration series based on top of this one. Adam Litke (1): blockjob: add API for async virDomainBlockJobAbort Eric Blake (4): blockjob: add qemu capabilities related to block pull jobs blockjob: optimize JSON event handler lookup blockjob: wire up qemu async virDomainBlockJobAbort blockjob: allow for fast-finishing job include/libvirt/libvirt.h.in | 10 +++ src/libvirt.c | 13 +++- src/qemu/qemu_capabilities.c | 3 + src/qemu/qemu_capabilities.h | 2 + src/qemu/qemu_driver.c | 106 ++++++++++++++++++++----- src/qemu/qemu_monitor.c | 11 ++- src/qemu/qemu_monitor.h | 8 +- src/qemu/qemu_monitor_json.c | 179 +++++++++++++++++++++++++++--------------- src/qemu/qemu_monitor_json.h | 4 +- tools/virsh.c | 51 ++++++++----- tools/virsh.pod | 35 ++++++--- 11 files changed, 300 insertions(+), 122 deletions(-) -- 1.7.7.6

2 14

[libvirt] [PATCH] qemu: optimize JSON event handler lookup
by Eric Blake 11 Apr '12

11 Apr '12

Probably in the noise, but this will let us scale more efficiently as we recognize ever more qemu events. * src/qemu/qemu_monitor_json.c (eventHandlers): Sort. (eventSearch): New helper function. (qemuMonitorJSONIOProcessEvent): Optimize event lookup. --- In reply to: https://www.redhat.com/archives/libvir-list/2012-April/msg00416.html src/qemu/qemu_monitor_json.c | 55 +++++++++++++++++++++++------------------ 1 files changed, 31 insertions(+), 24 deletions(-) diff --git a/src/qemu/qemu_monitor_json.c b/src/qemu/qemu_monitor_json.c index 7c893d4..d142517 100644 --- a/src/qemu/qemu_monitor_json.c +++ b/src/qemu/qemu_monitor_json.c @@ -67,37 +67,45 @@ static void qemuMonitorJSONHandlePMSuspend(qemuMonitorPtr mon, virJSONValuePtr d static void qemuMonitorJSONHandleBlockJobCompleted(qemuMonitorPtr mon, virJSONValuePtr data); static void qemuMonitorJSONHandleBlockJobCanceled(qemuMonitorPtr mon, virJSONValuePtr data); -static struct { +typedef struct { const char *type; void (*handler)(qemuMonitorPtr mon, virJSONValuePtr data); -} eventHandlers[] = { - { "SHUTDOWN", qemuMonitorJSONHandleShutdown, }, - { "RESET", qemuMonitorJSONHandleReset, }, +} qemuEventHandler; +static qemuEventHandler eventHandlers[] = { + { "BLOCK_IO_ERROR", qemuMonitorJSONHandleIOError, }, + { "BLOCK_JOB_CANCELLED", qemuMonitorJSONHandleBlockJobCanceled, }, + { "BLOCK_JOB_COMPLETED", qemuMonitorJSONHandleBlockJobCompleted, }, + { "DEVICE_TRAY_MOVED", qemuMonitorJSONHandleTrayChange, }, { "POWERDOWN", qemuMonitorJSONHandlePowerdown, }, - { "STOP", qemuMonitorJSONHandleStop, }, + { "RESET", qemuMonitorJSONHandleReset, }, { "RTC_CHANGE", qemuMonitorJSONHandleRTCChange, }, - { "WATCHDOG", qemuMonitorJSONHandleWatchdog, }, - { "BLOCK_IO_ERROR", qemuMonitorJSONHandleIOError, }, - { "VNC_CONNECTED", qemuMonitorJSONHandleVNCConnect, }, - { "VNC_INITIALIZED", qemuMonitorJSONHandleVNCInitialize, }, - { "VNC_DISCONNECTED", qemuMonitorJSONHandleVNCDisconnect, }, + { "SHUTDOWN", qemuMonitorJSONHandleShutdown, }, { "SPICE_CONNECTED", qemuMonitorJSONHandleSPICEConnect, }, - { "SPICE_INITIALIZED", qemuMonitorJSONHandleSPICEInitialize, }, { "SPICE_DISCONNECTED", qemuMonitorJSONHandleSPICEDisconnect, }, - { "DEVICE_TRAY_MOVED", qemuMonitorJSONHandleTrayChange, }, - { "WAKEUP", qemuMonitorJSONHandlePMWakeup, }, + { "SPICE_INITIALIZED", qemuMonitorJSONHandleSPICEInitialize, }, + { "STOP", qemuMonitorJSONHandleStop, }, { "SUSPEND", qemuMonitorJSONHandlePMSuspend, }, - { "BLOCK_JOB_COMPLETED", qemuMonitorJSONHandleBlockJobCompleted, }, - { "BLOCK_JOB_CANCELLED", qemuMonitorJSONHandleBlockJobCanceled, }, + { "VNC_CONNECTED", qemuMonitorJSONHandleVNCConnect, }, + { "VNC_DISCONNECTED", qemuMonitorJSONHandleVNCDisconnect, }, + { "VNC_INITIALIZED", qemuMonitorJSONHandleVNCInitialize, }, + { "WAKEUP", qemuMonitorJSONHandlePMWakeup, }, + { "WATCHDOG", qemuMonitorJSONHandleWatchdog, }, + /* We use bsearch, so keep this list sorted. */ }; +static int +eventSearch(const void *key, const void *elt) { + const char *type = key; + const qemuEventHandler *handler = elt; + return strcmp(type, handler->type); +} static int qemuMonitorJSONIOProcessEvent(qemuMonitorPtr mon, virJSONValuePtr obj) { const char *type; - int i; + qemuEventHandler *handler; VIR_DEBUG("mon=%p obj=%p", mon, obj); type = virJSONValueObjectGetString(obj, "event"); @@ -107,14 +115,13 @@ qemuMonitorJSONIOProcessEvent(qemuMonitorPtr mon, return -1; } - for (i = 0 ; i < ARRAY_CARDINALITY(eventHandlers) ; i++) { - if (STREQ(eventHandlers[i].type, type)) { - virJSONValuePtr data = virJSONValueObjectGet(obj, "data"); - VIR_DEBUG("handle %s handler=%p data=%p", type, - eventHandlers[i].handler, data); - (eventHandlers[i].handler)(mon, data); - break; - } + handler = bsearch(type, eventHandlers, ARRAY_CARDINALITY(eventHandlers), + sizeof(eventHandlers[0]), eventSearch); + if (handler) { + virJSONValuePtr data = virJSONValueObjectGet(obj, "data"); + VIR_DEBUG("handle %s handler=%p data=%p", type, + handler->handler, data); + (handler->handler)(mon, data); } return 0; } -- 1.7.7.6

2 2

[libvirt] [libvirt-glib] libvirt-gconfig: Move symbols introduced post 0.0.4 to the correct versions
by Guido Günther 11 Apr '12

11 Apr '12

Triggered by http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=667683 --- If this looks correct I'll have a look at libvirt-glib and libvirt-gobject too. -- Guido libvirt-gconfig/libvirt-gconfig.sym | 50 +++++++++++++++++++++++------------ 1 files changed, 33 insertions(+), 17 deletions(-) diff --git a/libvirt-gconfig/libvirt-gconfig.sym b/libvirt-gconfig/libvirt-gconfig.sym index 717c3c9..405ad98 100644 --- a/libvirt-gconfig/libvirt-gconfig.sym +++ b/libvirt-gconfig/libvirt-gconfig.sym @@ -14,10 +14,6 @@ LIBVIRT_GCONFIG_0.0.4 { gvir_config_domain_new; gvir_config_domain_new_from_xml; gvir_config_domain_set_clock; - gvir_config_domain_set_custom_xml; - gvir_config_domain_get_custom_xml; - gvir_config_domain_get_description; - gvir_config_domain_set_description; gvir_config_domain_get_devices; gvir_config_domain_set_devices; gvir_config_domain_get_features; @@ -51,10 +47,6 @@ LIBVIRT_GCONFIG_0.0.4 { gvir_config_domain_chardev_source_pty_new_from_xml; gvir_config_domain_source_pty_set_path; - gvir_config_domain_chardev_source_spicevmc_get_type; - gvir_config_domain_chardev_source_spicevmc_new; - gvir_config_domain_chardev_source_spicevmc_new_from_xml; - gvir_config_domain_clock_get_type; gvir_config_domain_clock_offset_get_type; gvir_config_domain_clock_new; @@ -79,8 +71,6 @@ LIBVIRT_GCONFIG_0.0.4 { gvir_config_domain_disk_type_get_type; gvir_config_domain_disk_new; gvir_config_domain_disk_new_from_xml; - gvir_config_domain_disk_get_driver_cache; - gvir_config_domain_disk_set_driver_cache; gvir_config_domain_disk_get_driver_name; gvir_config_domain_disk_set_driver_name; gvir_config_domain_disk_get_driver_type; @@ -96,7 +86,6 @@ LIBVIRT_GCONFIG_0.0.4 { gvir_config_domain_disk_get_target_dev; gvir_config_domain_disk_set_target_dev; gvir_config_domain_disk_get_disk_type; - gvir_config_domain_disk_set_readonly; gvir_config_domain_disk_set_type; gvir_config_domain_filesys_get_type; @@ -146,13 +135,9 @@ LIBVIRT_GCONFIG_0.0.4 { gvir_config_domain_interface_get_type; gvir_config_domain_interface_link_state_get_type; gvir_config_domain_interface_set_ifname; - gvir_config_domain_interface_get_ifname; gvir_config_domain_interface_set_link_state; - gvir_config_domain_interface_get_link_state; gvir_config_domain_interface_set_mac; - gvir_config_domain_interface_get_mac; gvir_config_domain_interface_set_model; - gvir_config_domain_interface_get_model; gvir_config_domain_interface_bridge_get_type; gvir_config_domain_interface_bridge_new; @@ -327,8 +312,39 @@ LIBVIRT_GCONFIG_0.0.4 { gvir_config_xml_doc_get_type; gvir_config_xml_doc_new; - local: - *; }; +LIBVIRT_GCONFIG_0.0.5 { + global: + gvir_config_domain_disk_get_driver_cache; + gvir_config_domain_disk_set_driver_cache; + + gvir_config_domain_disk_set_readonly; + + gvir_config_domain_set_custom_xml; + gvir_config_domain_get_custom_xml; + gvir_config_domain_get_description; + gvir_config_domain_set_description; +} LIBVIRT_GCONFIG_0.0.4; + +LIBVIRT_GCONFIG_0.0.6 { + global: + gvir_config_domain_interface_get_ifname; + gvir_config_domain_interface_get_link_state; + gvir_config_domain_interface_get_mac; + gvir_config_domain_interface_get_model; +} LIBVIRT_GCONFIG_0.0.5; + +LIBVIRT_GCONFIG_0.0.7 { + global: + gvir_config_domain_chardev_source_spicevmc_get_type; + gvir_config_domain_chardev_source_spicevmc_new; + gvir_config_domain_chardev_source_spicevmc_new_from_xml; +} LIBVIRT_GCONFIG_0.0.6; + # .... define new API here using predicted next version number .... + +LIBVIRT_GCONFIG_PRIVATE_0.0.7 { + local: + *; +}; -- 1.7.9.1

3 9

[libvirt] USB redirection support
by Christophe Fergeau 11 Apr '12

11 Apr '12

Hey, This patch series adds the needed classes for apps to be able to use USB redirection. I've followed http://hansdegoede.livejournal.com/11084.html and added all the API needed to be able to get the same XML. However, I'm not 100% sure all these attributes are required, in particular the various addresses. Given that for now libvirt-gconfig users tend to reuse what test-domain-create does, it would be useful to know which parts are required and which are not to get working USB redirection. Thanks, Christophe

2 15

[libvirt] [PATCH] qemu: Fix deadlock when qemuDomainOpenConsole cleans up a connection
by Peter Krempa 11 Apr '12

11 Apr '12

The new safe console handling introduced a possibility to deadlock the qemu driver when a new console connection forcibly disconnects a previous console stream that belongs to an already closed connection. The virStreamFree function calls subsequently a the virReleaseConnect function that tries to lock the driver while discarding the connection, but the driver was already locked in qemuDomainOpenConsole. Backtrace of the deadlocked thread: 0 0x00007f66e5aa7f14 in __lll_lock_wait () from /lib64/libpthread.so.0 1 0x00007f66e5aa3411 in _L_lock_500 () from /lib64/libpthread.so.0 2 0x00007f66e5aa322a in pthread_mutex_lock () from/lib64/libpthread.so.0 3 0x0000000000462bbd in qemudClose () 4 0x00007f66e6e178eb in virReleaseConnect () from/usr/lib64/libvirt.so.0 5 0x00007f66e6e19c8c in virUnrefStream () from /usr/lib64/libvirt.so.0 6 0x00007f66e6e3d1de in virStreamFree () from /usr/lib64/libvirt.so.0 7 0x00007f66e6e09a5d in virConsoleHashEntryFree () from/usr/lib64/libvirt.so.0 8 0x00007f66e6db7282 in virHashRemoveEntry () from/usr/lib64/libvirt.so.0 9 0x00007f66e6e09c4e in virConsoleOpen () from /usr/lib64/libvirt.so.0 10 0x00000000004526e9 in qemuDomainOpenConsole () 11 0x00007f66e6e421f1 in virDomainOpenConsole () from/usr/lib64/libvirt.so.0 12 0x00000000004361e4 in remoteDispatchDomainOpenConsoleHelper () 13 0x00007f66e6e80375 in virNetServerProgramDispatch () from/usr/lib64/libvirt.so.0 14 0x00007f66e6e7ae11 in virNetServerHandleJob () from/usr/lib64/libvirt.so.0 15 0x00007f66e6da897d in virThreadPoolWorker () from/usr/lib64/libvirt.so.0 16 0x00007f66e6da7ff6 in virThreadHelper () from/usr/lib64/libvirt.so.0 17 0x00007f66e5aa0c5c in start_thread () from /lib64/libpthread.so.0 18 0x00007f66e57e7fcd in clone () from /lib64/libc.so.6 * src/qemu/qemu_driver.c: qemuDomainOpenConsole() -- unlock the qemu driver right after acquiring the domain object --- Found while writing tests for the libvirt-test-API, its 100% reproducible: 1: Start a domain with serial console and run "cat" on this console 2: download python reproducer: http://files.pipo.sk/console_deadlock.py 3: set guest name in the script 4: run the reproducer and "virsh console 'guestname' --force" after that 5: <deadlocked> src/qemu/qemu_driver.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index d9e35be..e18e72d 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -11494,6 +11494,7 @@ qemuDomainOpenConsole(virDomainPtr dom, qemuDriverLock(driver); virUUIDFormat(dom->uuid, uuidstr); vm = virDomainFindByUUID(&driver->domains, dom->uuid); + qemuDriverUnlock(driver); if (!vm) { qemuReportError(VIR_ERR_NO_DOMAIN, _("no domain with matching uuid '%s'"), uuidstr); @@ -11558,7 +11559,6 @@ qemuDomainOpenConsole(virDomainPtr dom, cleanup: if (vm) virDomainObjUnlock(vm); - qemuDriverUnlock(driver); return ret; } -- 1.7.3.4

2 2

[libvirt] Etsy uses Libvirt... :)
by Justin Clift 10 Apr '12

10 Apr '12

Hey everyone, Just noticed that Etsy uses libvirt: http://codeascraft.etsy.com/2012/03/13/making-it-virtually-easy-to-deploy-o… Of specific note, there's this bit: "Libvirt supports live migrations across non-shared storage (in QEMU 0.12.2+) with zero downtime which makes it easy to allocate and balance VM’s across hosts if adjustments need to be made throughout the pool." Seems like a win, eh? :) Regards and best wishes, Justin Clift -- Aeolus Community Manager http://www.aeolusproject.org

2 3

[libvirt] [PATCH 1/2] qemu: Track job owner for better debugging
by Jiri Denemark 10 Apr '12

10 Apr '12

In case an API fails with "cannot acquire state change lock", searching for the API that possibly forgot to end its job is not always easy. Let's keep track of the job owner and print it out for easier identification. --- src/qemu/qemu_domain.c | 53 +++++++++++++++++++++++++++++++++++++++++++++ src/qemu/qemu_domain.h | 4 +++ src/qemu/qemu_migration.c | 1 + src/qemu/qemu_process.c | 6 +--- 4 files changed, 60 insertions(+), 4 deletions(-) diff --git a/src/qemu/qemu_domain.c b/src/qemu/qemu_domain.c index da1f57f..7f1f8ee 100644 --- a/src/qemu/qemu_domain.c +++ b/src/qemu/qemu_domain.c @@ -147,6 +147,7 @@ qemuDomainObjResetJob(qemuDomainObjPrivatePtr priv) struct qemuDomainJobObj *job = &priv->job; job->active = QEMU_JOB_NONE; + job->owner = 0; } static void @@ -155,6 +156,7 @@ qemuDomainObjResetAsyncJob(qemuDomainObjPrivatePtr priv) struct qemuDomainJobObj *job = &priv->job; job->asyncJob = QEMU_ASYNC_JOB_NONE; + job->asyncOwner = 0; job->phase = 0; job->mask = DEFAULT_JOB_MASK; job->start = 0; @@ -169,13 +171,25 @@ qemuDomainObjRestoreJob(virDomainObjPtr obj, memset(job, 0, sizeof(*job)); job->active = priv->job.active; + job->owner = priv->job.owner; job->asyncJob = priv->job.asyncJob; + job->asyncOwner = priv->job.asyncOwner; job->phase = priv->job.phase; qemuDomainObjResetJob(priv); qemuDomainObjResetAsyncJob(priv); } +void +qemuDomainObjTransferJob(virDomainObjPtr obj) +{ + qemuDomainObjPrivatePtr priv = obj->privateData; + + VIR_DEBUG("Changing job owner from %d to %d", + priv->job.owner, virThreadSelfID()); + priv->job.owner = virThreadSelfID(); +} + static void qemuDomainObjFreeJob(qemuDomainObjPrivatePtr priv) { @@ -664,11 +678,23 @@ qemuDomainObjSetJobPhase(struct qemud_driver *driver, int phase) { qemuDomainObjPrivatePtr priv = obj->privateData; + int me = virThreadSelfID(); if (!priv->job.asyncJob) return; + VIR_DEBUG("Setting '%s' phase to '%s'", + qemuDomainAsyncJobTypeToString(priv->job.asyncJob), + qemuDomainAsyncJobPhaseToString(priv->job.asyncJob, phase)); + + if (priv->job.asyncOwner && me != priv->job.asyncOwner) { + VIR_WARN("'%s' async job is owned by thread %d", + qemuDomainAsyncJobTypeToString(priv->job.asyncJob), + priv->job.asyncOwner); + } + priv->job.phase = phase; + priv->job.asyncOwner = me; qemuDomainObjSaveJob(driver, obj); } @@ -695,6 +721,22 @@ qemuDomainObjDiscardAsyncJob(struct qemud_driver *driver, virDomainObjPtr obj) qemuDomainObjSaveJob(driver, obj); } +void +qemuDomainObjReleaseAsyncJob(virDomainObjPtr obj) +{ + qemuDomainObjPrivatePtr priv = obj->privateData; + + VIR_DEBUG("Releasing ownership of '%s' async job", + qemuDomainAsyncJobTypeToString(priv->job.asyncJob)); + + if (priv->job.asyncOwner != virThreadSelfID()) { + VIR_WARN("'%s' async job is owned by thread %d", + qemuDomainAsyncJobTypeToString(priv->job.asyncJob), + priv->job.asyncOwner); + } + priv->job.asyncOwner = 0; +} + static bool qemuDomainNestedJobAllowed(qemuDomainObjPrivatePtr priv, enum qemuDomainJob job) { @@ -764,11 +806,13 @@ retry: qemuDomainJobTypeToString(job), qemuDomainAsyncJobTypeToString(priv->job.asyncJob)); priv->job.active = job; + priv->job.owner = virThreadSelfID(); } else { VIR_DEBUG("Starting async job: %s", qemuDomainAsyncJobTypeToString(asyncJob)); qemuDomainObjResetAsyncJob(priv); priv->job.asyncJob = asyncJob; + priv->job.asyncOwner = virThreadSelfID(); priv->job.start = now; } @@ -784,6 +828,15 @@ retry: return 0; error: + VIR_WARN("Cannot start job (%s, %s) for domain %s;" + " current job is (%s, %s) owned by (%d, %d)", + qemuDomainJobTypeToString(job), + qemuDomainAsyncJobTypeToString(asyncJob), + obj->def->name, + qemuDomainJobTypeToString(priv->job.active), + qemuDomainAsyncJobTypeToString(priv->job.asyncJob), + priv->job.owner, priv->job.asyncOwner); + if (errno == ETIMEDOUT) qemuReportError(VIR_ERR_OPERATION_TIMEOUT, "%s", _("cannot acquire state change lock")); diff --git a/src/qemu/qemu_domain.h b/src/qemu/qemu_domain.h index b3eecd3..ce52569 100644 --- a/src/qemu/qemu_domain.h +++ b/src/qemu/qemu_domain.h @@ -96,9 +96,11 @@ VIR_ENUM_DECL(qemuDomainAsyncJob) struct qemuDomainJobObj { virCond cond; /* Use to coordinate jobs */ enum qemuDomainJob active; /* Currently running job */ + int owner; /* Thread which set current job */ virCond asyncCond; /* Use to coordinate with async jobs */ enum qemuDomainAsyncJob asyncJob; /* Currently active async job */ + int asyncOwner; /* Thread which set current async job */ int phase; /* Job phase (mainly for migrations) */ unsigned long long mask; /* Jobs allowed during async job */ unsigned long long start; /* When the async job started */ @@ -203,8 +205,10 @@ void qemuDomainObjSetAsyncJobMask(virDomainObjPtr obj, unsigned long long allowedJobs); void qemuDomainObjRestoreJob(virDomainObjPtr obj, struct qemuDomainJobObj *job); +void qemuDomainObjTransferJob(virDomainObjPtr obj); void qemuDomainObjDiscardAsyncJob(struct qemud_driver *driver, virDomainObjPtr obj); +void qemuDomainObjReleaseAsyncJob(virDomainObjPtr obj); void qemuDomainObjEnterMonitor(struct qemud_driver *driver, virDomainObjPtr obj) diff --git a/src/qemu/qemu_migration.c b/src/qemu/qemu_migration.c index ff6f273..c1bb93a 100644 --- a/src/qemu/qemu_migration.c +++ b/src/qemu/qemu_migration.c @@ -3242,6 +3242,7 @@ qemuMigrationJobStartPhase(struct qemud_driver *driver, int qemuMigrationJobContinue(virDomainObjPtr vm) { + qemuDomainObjReleaseAsyncJob(vm); return virDomainObjUnref(vm); } diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index 96f39e8..19569cf 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -3038,8 +3038,8 @@ qemuProcessReconnect(void *opaque) priv = obj->privateData; - /* Set fake job so that EnterMonitor* doesn't want to start a new one */ - priv->job.active = QEMU_JOB_MODIFY; + /* Job was started by the caller for us */ + qemuDomainObjTransferJob(obj); /* Hold an extra reference because we can't allow 'vm' to be * deleted if qemuConnectMonitor() failed */ @@ -3119,8 +3119,6 @@ qemuProcessReconnect(void *opaque) if (qemuProcessRecoverJob(driver, obj, conn, &oldjob) < 0) goto error; - priv->job.active = QEMU_JOB_NONE; - /* update domain state XML with possibly updated state in virDomainObj */ if (virDomainSaveStatus(driver->caps, driver->stateDir, obj) < 0) goto error; -- 1.7.8.5

2 2

[libvirt] [PATCH] qemu: Avoid excessive calls to qemuDomainObjSaveJob()
by Jiri Denemark 10 Apr '12

10 Apr '12

As reported by Daniel Berrangé, we have a huge performance regression for virDomainGetInfo() due to the change which makes virDomainEndJob() save the XML status file every time it is called. Previous to that change, 2000 calls to virDomainGetInfo() took ~2.5 seconds. After that change, 2000 calls to virDomainGetInfo() take 2 *minutes* 45 secs. We made the change to be able to recover from libvirtd restart in the middle of a job. However, only destroy and async jobs are taken care of. Thus it makes more sense to only save domain state XML when these jobs are started/stopped. --- src/qemu/qemu_domain.c | 21 ++++++++++++++++++--- src/qemu/qemu_domain.h | 5 +++++ src/qemu/qemu_process.c | 4 ++++ 3 files changed, 27 insertions(+), 3 deletions(-) diff --git a/src/qemu/qemu_domain.c b/src/qemu/qemu_domain.c index 69d9e6e..da1f57f 100644 --- a/src/qemu/qemu_domain.c +++ b/src/qemu/qemu_domain.c @@ -183,6 +183,12 @@ qemuDomainObjFreeJob(qemuDomainObjPrivatePtr priv) ignore_value(virCondDestroy(&priv->job.asyncCond)); } +static bool +qemuDomainTrackJob(enum qemuDomainJob job) +{ + return (QEMU_DOMAIN_TRACK_JOBS & JOB_MASK(job)) != 0; +} + static void *qemuDomainObjPrivateAlloc(void) { @@ -239,6 +245,7 @@ static int qemuDomainObjPrivateXMLFormat(virBufferPtr buf, void *data) { qemuDomainObjPrivatePtr priv = data; const char *monitorpath; + enum qemuDomainJob job; /* priv->monitor_chr is set only for qemu */ if (priv->monConfig) { @@ -284,6 +291,10 @@ static int qemuDomainObjPrivateXMLFormat(virBufferPtr buf, void *data) if (priv->lockState) virBufferAsprintf(buf, " <lockstate>%s</lockstate>\n", priv->lockState); + job = priv->job.active; + if (!qemuDomainTrackJob(job)) + priv->job.active = QEMU_JOB_NONE; + if (priv->job.active || priv->job.asyncJob) { virBufferAsprintf(buf, " <job type='%s' async='%s'", qemuDomainJobTypeToString(priv->job.active), @@ -295,6 +306,7 @@ static int qemuDomainObjPrivateXMLFormat(virBufferPtr buf, void *data) } virBufferAddLit(buf, "/>\n"); } + priv->job.active = job; if (priv->fakeReboot) virBufferAsprintf(buf, " <fakereboot/>\n"); @@ -766,7 +778,8 @@ retry: virDomainObjLock(obj); } - qemuDomainObjSaveJob(driver, obj); + if (qemuDomainTrackJob(job)) + qemuDomainObjSaveJob(driver, obj); return 0; @@ -862,15 +875,17 @@ int qemuDomainObjBeginAsyncJobWithDriver(struct qemud_driver *driver, int qemuDomainObjEndJob(struct qemud_driver *driver, virDomainObjPtr obj) { qemuDomainObjPrivatePtr priv = obj->privateData; + enum qemuDomainJob job = priv->job.active; priv->jobs_queued--; VIR_DEBUG("Stopping job: %s (async=%s)", - qemuDomainJobTypeToString(priv->job.active), + qemuDomainJobTypeToString(job), qemuDomainAsyncJobTypeToString(priv->job.asyncJob)); qemuDomainObjResetJob(priv); - qemuDomainObjSaveJob(driver, obj); + if (qemuDomainTrackJob(job)) + qemuDomainObjSaveJob(driver, obj); virCondSignal(&priv->job.cond); return virDomainObjUnref(obj); diff --git a/src/qemu/qemu_domain.h b/src/qemu/qemu_domain.h index adccfed..b3eecd3 100644 --- a/src/qemu/qemu_domain.h +++ b/src/qemu/qemu_domain.h @@ -53,6 +53,11 @@ JOB_MASK(QEMU_JOB_DESTROY) | \ JOB_MASK(QEMU_JOB_ABORT)) +/* Jobs which have to be tracked in domain state XML. */ +# define QEMU_DOMAIN_TRACK_JOBS \ + (JOB_MASK(QEMU_JOB_DESTROY) | \ + JOB_MASK(QEMU_JOB_ASYNC)) + /* Only 1 job is allowed at any time * A job includes *all* monitor commands, even those just querying * information, not merely actions */ diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index 481b4f3..96f39e8 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -2963,6 +2963,10 @@ qemuProcessRecoverJob(struct qemud_driver *driver, if (!virDomainObjIsActive(vm)) return -1; + /* In case any special handling is added for job type that has been ignored + * before, QEMU_DOMAIN_TRACK_JOBS (from qemu_domain.h) needs to be updated + * for the job to be properly tracked in domain state XML. + */ switch (job->active) { case QEMU_JOB_QUERY: /* harmless */ -- 1.7.8.5

2 1

[libvirt] [PATCH] build: avoid s390 compiler warnings
by Eric Blake 10 Apr '12

10 Apr '12

I noticed compiler warnings when building for the s390 architecture. * src/node_device/node_device_udev.c (udevDeviceMonitorStartup): Mark unused variable. * src/nodeinfo.c (linuxNodeInfoCPUPopulate): Avoid unused variable. --- Pushing under the trivial rule. src/node_device/node_device_udev.c | 4 ++-- src/nodeinfo.c | 3 ++- 2 files changed, 4 insertions(+), 3 deletions(-) diff --git a/src/node_device/node_device_udev.c b/src/node_device/node_device_udev.c index 438c2eb..6418015 100644 --- a/src/node_device/node_device_udev.c +++ b/src/node_device/node_device_udev.c @@ -1,7 +1,7 @@ /* * node_device_udev.c: node device enumeration - libudev implementation * - * Copyright (C) 2009-2011 Red Hat, Inc. + * Copyright (C) 2009-2012 Red Hat, Inc. * * This library is free software; you can redistribute it and/or * modify it under the terms of the GNU Lesser General Public @@ -1604,7 +1604,7 @@ out: return ret; } -static int udevDeviceMonitorStartup(int privileged) +static int udevDeviceMonitorStartup(int privileged ATTRIBUTE_UNUSED) { udevPrivate *priv = NULL; struct udev *udev = NULL; diff --git a/src/nodeinfo.c b/src/nodeinfo.c index 26f8dc1..7e993a9 100644 --- a/src/nodeinfo.c +++ b/src/nodeinfo.c @@ -231,10 +231,10 @@ int linuxNodeInfoCPUPopulate(FILE *cpuinfo, /* NOTE: hyperthreads are ignored here; they are parsed out of /sys */ while (fgets(line, sizeof(line), cpuinfo) != NULL) { - char *buf = line; # if defined(__x86_64__) || \ defined(__amd64__) || \ defined(__i386__) + char *buf = line; if (STRPREFIX(buf, "cpu MHz")) { char *p; unsigned int ui; @@ -253,6 +253,7 @@ int linuxNodeInfoCPUPopulate(FILE *cpuinfo, } # elif defined(__powerpc__) || \ defined(__powerpc64__) + char *buf = line; if (STRPREFIX(buf, "clock")) { char *p; unsigned int ui; -- 1.7.7.6

1 0

[libvirt] [PATCH] daemon: Add libvirtd-config.c to the list of files to translate
by Jiri Denemark 10 Apr '12

10 Apr '12

--- po/POTFILES.in | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/po/POTFILES.in b/po/POTFILES.in index 5d5739a..4c49200 100644 --- a/po/POTFILES.in +++ b/po/POTFILES.in @@ -1,3 +1,4 @@ +daemon/libvirtd-config.c daemon/libvirtd.c daemon/qemu_dispatch.h daemon/remote.c -- 1.7.8.5

2 1

[libvirt] [PATCH] virsh: Clean up usage of boolean flag variables
by Peter Krempa 10 Apr '12

10 Apr '12

This patch cleans up variables used to store boolean command flags that are inquired by vshCommandOptBool to use the bool data type instead of an integer. Additionaly this patch cleans up flag variables that are inferred from existing flags. --- tools/virsh.c | 120 ++++++++++++++++++++++++++++---------------------------- 1 files changed, 60 insertions(+), 60 deletions(-) diff --git a/tools/virsh.c b/tools/virsh.c index cfdd040..1a2e191 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -933,9 +933,9 @@ static const vshCmdOptDef opts_list[] = { static bool cmdList(vshControl *ctl, const vshCmd *cmd ATTRIBUTE_UNUSED) { - int inactive = vshCommandOptBool(cmd, "inactive"); - int all = vshCommandOptBool(cmd, "all"); - int active = !inactive || all ? 1 : 0; + bool inactive = vshCommandOptBool(cmd, "inactive"); + bool all = vshCommandOptBool(cmd, "all"); + bool active = !inactive || all; int *ids = NULL, maxid = 0, i; char **names = NULL; int maxname = 0; @@ -1317,7 +1317,7 @@ cmdDomstate(vshControl *ctl, const vshCmd *cmd) { virDomainPtr dom; bool ret = true; - int showReason = vshCommandOptBool(cmd, "reason"); + bool showReason = vshCommandOptBool(cmd, "reason"); int state, reason; if (!vshConnectionUsability(ctl, ctl->conn)) @@ -1966,9 +1966,9 @@ cmdDomIftune(vshControl *ctl, const vshCmd *cmd) int nparams = 0; virTypedParameterPtr params = NULL; bool ret = false; - int current = vshCommandOptBool(cmd, "current"); - int config = vshCommandOptBool(cmd, "config"); - int live = vshCommandOptBool(cmd, "live"); + bool current = vshCommandOptBool(cmd, "current"); + bool config = vshCommandOptBool(cmd, "config"); + bool live = vshCommandOptBool(cmd, "live"); virNetDevBandwidthRate inbound, outbound; int i; @@ -2626,7 +2626,7 @@ cmdCreate(vshControl *ctl, const vshCmd *cmd) bool ret = true; char *buffer; #ifndef WIN32 - int console = vshCommandOptBool(cmd, "console"); + bool console = vshCommandOptBool(cmd, "console"); #endif unsigned int flags = VIR_DOMAIN_NONE; @@ -3094,7 +3094,7 @@ cmdStart(vshControl *ctl, const vshCmd *cmd) virDomainPtr dom; bool ret = false; #ifndef WIN32 - int console = vshCommandOptBool(cmd, "console"); + bool console = vshCommandOptBool(cmd, "console"); #endif unsigned int flags = VIR_DOMAIN_NONE; int rc; @@ -3786,9 +3786,9 @@ cmdSchedinfo(vshControl *ctl, const vshCmd *cmd) int i, ret; bool ret_val = false; unsigned int flags = 0; - int current = vshCommandOptBool(cmd, "current"); - int config = vshCommandOptBool(cmd, "config"); - int live = vshCommandOptBool(cmd, "live"); + bool current = vshCommandOptBool(cmd, "current"); + bool config = vshCommandOptBool(cmd, "config"); + bool live = vshCommandOptBool(cmd, "live"); if (current) { if (live || config) { @@ -4921,11 +4921,11 @@ cmdVcpucount(vshControl *ctl, const vshCmd *cmd) { virDomainPtr dom; bool ret = true; - int maximum = vshCommandOptBool(cmd, "maximum"); - int active = vshCommandOptBool(cmd, "active"); - int config = vshCommandOptBool(cmd, "config"); - int live = vshCommandOptBool(cmd, "live"); - int current = vshCommandOptBool(cmd, "current"); + bool maximum = vshCommandOptBool(cmd, "maximum"); + bool active = vshCommandOptBool(cmd, "active"); + bool config = vshCommandOptBool(cmd, "config"); + bool live = vshCommandOptBool(cmd, "live"); + bool current = vshCommandOptBool(cmd, "current"); bool all = maximum + active + current + config + live == 0; int count; @@ -5246,9 +5246,9 @@ cmdVcpuPin(vshControl *ctl, const vshCmd *cmd) int i, cpu, lastcpu, maxcpu, ncpus; bool unuse = false; const char *cur; - int config = vshCommandOptBool(cmd, "config"); - int live = vshCommandOptBool(cmd, "live"); - int current = vshCommandOptBool(cmd, "current"); + bool config = vshCommandOptBool(cmd, "config"); + bool live = vshCommandOptBool(cmd, "live"); + bool current = vshCommandOptBool(cmd, "current"); bool query = false; /* Query mode if no cpulist */ unsigned int flags = 0; @@ -5479,10 +5479,10 @@ cmdSetvcpus(vshControl *ctl, const vshCmd *cmd) virDomainPtr dom; int count = 0; bool ret = true; - int maximum = vshCommandOptBool(cmd, "maximum"); - int config = vshCommandOptBool(cmd, "config"); - int live = vshCommandOptBool(cmd, "live"); - int current = vshCommandOptBool(cmd, "current"); + bool maximum = vshCommandOptBool(cmd, "maximum"); + bool config = vshCommandOptBool(cmd, "config"); + bool live = vshCommandOptBool(cmd, "live"); + bool current = vshCommandOptBool(cmd, "current"); unsigned int flags = 0; if (current) { @@ -5853,9 +5853,9 @@ cmdSetmem(vshControl *ctl, const vshCmd *cmd) unsigned long long max; unsigned long kibibytes = 0; bool ret = true; - int config = vshCommandOptBool(cmd, "config"); - int live = vshCommandOptBool(cmd, "live"); - int current = vshCommandOptBool(cmd, "current"); + bool config = vshCommandOptBool(cmd, "config"); + bool live = vshCommandOptBool(cmd, "live"); + bool current = vshCommandOptBool(cmd, "current"); unsigned int flags = 0; if (current) { @@ -5935,9 +5935,9 @@ cmdSetmaxmem(vshControl *ctl, const vshCmd *cmd) unsigned long long max; unsigned long kibibytes = 0; bool ret = true; - int config = vshCommandOptBool(cmd, "config"); - int live = vshCommandOptBool(cmd, "live"); - int current = vshCommandOptBool(cmd, "current"); + bool config = vshCommandOptBool(cmd, "config"); + bool live = vshCommandOptBool(cmd, "live"); + bool current = vshCommandOptBool(cmd, "current"); unsigned int flags = VIR_DOMAIN_MEM_MAXIMUM; if (current) { @@ -6026,9 +6026,9 @@ cmdBlkiotune(vshControl * ctl, const vshCmd * cmd) virTypedParameterPtr params = NULL, temp = NULL; bool ret = false; unsigned int flags = 0; - int current = vshCommandOptBool(cmd, "current"); - int config = vshCommandOptBool(cmd, "config"); - int live = vshCommandOptBool(cmd, "live"); + bool current = vshCommandOptBool(cmd, "current"); + bool config = vshCommandOptBool(cmd, "config"); + bool live = vshCommandOptBool(cmd, "live"); if (current) { if (live || config) { @@ -6201,9 +6201,9 @@ cmdMemtune(vshControl *ctl, const vshCmd *cmd) virTypedParameterPtr params = NULL, temp = NULL; bool ret = false; unsigned int flags = 0; - int current = vshCommandOptBool(cmd, "current"); - int config = vshCommandOptBool(cmd, "config"); - int live = vshCommandOptBool(cmd, "live"); + bool current = vshCommandOptBool(cmd, "current"); + bool config = vshCommandOptBool(cmd, "config"); + bool live = vshCommandOptBool(cmd, "live"); if (current) { if (live || config) { @@ -6365,9 +6365,9 @@ cmdNumatune(vshControl * ctl, const vshCmd * cmd) const char *nodeset = NULL; bool ret = false; unsigned int flags = 0; - int current = vshCommandOptBool(cmd, "current"); - int config = vshCommandOptBool(cmd, "config"); - int live = vshCommandOptBool(cmd, "live"); + bool current = vshCommandOptBool(cmd, "current"); + bool config = vshCommandOptBool(cmd, "config"); + bool live = vshCommandOptBool(cmd, "live"); const char *mode = NULL; if (current) { @@ -6831,9 +6831,9 @@ cmdDumpXML(vshControl *ctl, const vshCmd *cmd) bool ret = true; char *dump; unsigned int flags = 0; - int inactive = vshCommandOptBool(cmd, "inactive"); - int secure = vshCommandOptBool(cmd, "security-info"); - int update = vshCommandOptBool(cmd, "update-cpu"); + bool inactive = vshCommandOptBool(cmd, "inactive"); + bool secure = vshCommandOptBool(cmd, "security-info"); + bool update = vshCommandOptBool(cmd, "update-cpu"); if (inactive) flags |= VIR_DOMAIN_XML_INACTIVE; @@ -7800,9 +7800,9 @@ cmdBlkdeviotune(vshControl *ctl, const vshCmd *cmd) virTypedParameterPtr params = NULL, temp = NULL; unsigned int flags = 0, i = 0; int rv = 0; - int current = vshCommandOptBool(cmd, "current"); - int config = vshCommandOptBool(cmd, "config"); - int live = vshCommandOptBool(cmd, "live"); + bool current = vshCommandOptBool(cmd, "current"); + bool config = vshCommandOptBool(cmd, "config"); + bool live = vshCommandOptBool(cmd, "live"); bool ret = false; if (current) { @@ -8336,9 +8336,9 @@ static const vshCmdOptDef opts_network_list[] = { static bool cmdNetworkList(vshControl *ctl, const vshCmd *cmd ATTRIBUTE_UNUSED) { - int inactive = vshCommandOptBool(cmd, "inactive"); - int all = vshCommandOptBool(cmd, "all"); - int active = !inactive || all ? 1 : 0; + bool inactive = vshCommandOptBool(cmd, "inactive"); + bool all = vshCommandOptBool(cmd, "all"); + bool active = !inactive || all; int maxactive = 0, maxinactive = 0, i; char **activeNames = NULL, **inactiveNames = NULL; inactive |= all; @@ -8609,9 +8609,9 @@ static const vshCmdOptDef opts_interface_list[] = { static bool cmdInterfaceList(vshControl *ctl, const vshCmd *cmd ATTRIBUTE_UNUSED) { - int inactive = vshCommandOptBool(cmd, "inactive"); - int all = vshCommandOptBool(cmd, "all"); - int active = !inactive || all ? 1 : 0; + bool inactive = vshCommandOptBool(cmd, "inactive"); + bool all = vshCommandOptBool(cmd, "all"); + bool active = !inactive || all; int maxactive = 0, maxinactive = 0, i; char **activeNames = NULL, **inactiveNames = NULL; inactive |= all; @@ -8786,7 +8786,7 @@ cmdInterfaceDumpXML(vshControl *ctl, const vshCmd *cmd) bool ret = true; char *dump; unsigned int flags = 0; - int inactive = vshCommandOptBool(cmd, "inactive"); + bool inactive = vshCommandOptBool(cmd, "inactive"); if (inactive) flags |= VIR_INTERFACE_XML_INACTIVE; @@ -10038,7 +10038,7 @@ cmdPoolCreateAs(vshControl *ctl, const vshCmd *cmd) virStoragePoolPtr pool; const char *name; char *xml; - int printXML = vshCommandOptBool(cmd, "print-xml"); + bool printXML = vshCommandOptBool(cmd, "print-xml"); if (!vshConnectionUsability(ctl, ctl->conn)) return false; @@ -10126,7 +10126,7 @@ cmdPoolDefineAs(vshControl *ctl, const vshCmd *cmd) virStoragePoolPtr pool; const char *name; char *xml; - int printXML = vshCommandOptBool(cmd, "print-xml"); + bool printXML = vshCommandOptBool(cmd, "print-xml"); if (!vshConnectionUsability(ctl, ctl->conn)) return false; @@ -10400,10 +10400,10 @@ cmdPoolList(vshControl *ctl, const vshCmd *cmd ATTRIBUTE_UNUSED) struct poolInfoText *poolInfoTexts = NULL; /* Determine the options passed by the user */ - int all = vshCommandOptBool(cmd, "all"); - int details = vshCommandOptBool(cmd, "details"); - int inactive = vshCommandOptBool(cmd, "inactive"); - int active = !inactive || all ? 1 : 0; + bool all = vshCommandOptBool(cmd, "all"); + bool details = vshCommandOptBool(cmd, "details"); + bool inactive = vshCommandOptBool(cmd, "inactive"); + bool active = !inactive || all; inactive |= all; /* Check the connection to libvirtd daemon is still working */ @@ -12062,7 +12062,7 @@ cmdVolList(vshControl *ctl, const vshCmd *cmd ATTRIBUTE_UNUSED) char *outputStr = NULL; const char *unit; double val; - int details = vshCommandOptBool(cmd, "details"); + bool details = vshCommandOptBool(cmd, "details"); int numVolumes = 0, i; int ret; bool functionReturn; @@ -13006,7 +13006,7 @@ cmdNodeListDevices (vshControl *ctl, const vshCmd *cmd ATTRIBUTE_UNUSED) const char *cap = NULL; char **devices; int num_devices, i; - int tree = vshCommandOptBool(cmd, "tree"); + bool tree = vshCommandOptBool(cmd, "tree"); if (!vshConnectionUsability(ctl, ctl->conn)) return false; -- 1.7.3.4

2 2

[libvirt] [PATCH] virsh: Clarify use of the --managed-save flag for the list command
by Peter Krempa 10 Apr '12

10 Apr '12

The documentation for the flag doesn't clearly state that the flag only enhances the output and the user needs to specify other flags to list inactive domains, that are enhanced by this flag. --- tools/virsh.c | 2 +- tools/virsh.pod | 5 +++-- 2 files changed, 4 insertions(+), 3 deletions(-) diff --git a/tools/virsh.c b/tools/virsh.c index cfdd040..7477d32 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -924,7 +924,7 @@ static const vshCmdOptDef opts_list[] = { {"name", VSH_OT_BOOL, 0, N_("list domain names only")}, {"table", VSH_OT_BOOL, 0, N_("list table (default)")}, {"managed-save", VSH_OT_BOOL, 0, - N_("mark domains with managed save state")}, + N_("mark inactive domains with managed save state")}, {"title", VSH_OT_BOOL, 0, N_("show short domain description")}, {NULL, 0, 0, NULL} }; diff --git a/tools/virsh.pod b/tools/virsh.pod index a60e667..c6e0ff3 100644 --- a/tools/virsh.pod +++ b/tools/virsh.pod @@ -381,8 +381,9 @@ into s3 state. =back If I<--managed-save> is specified, then domains that have managed save -state (only possible if they are in the B<shut off> state) will -instead show as B<saved> in the listing. This flag is usable only with the +state (only possible if they are in the B<shut off> state, so you need to +specify I<--inactive> or I<--all> to actually list them) will instead +show as B<saved> in the listing. This flag is usable only with the default I<--table> output. If I<--name> is specified, domain names are printed instead of the table -- 1.7.3.4

2 2

[libvirt] [RESEND] libvirt: xen: Shutdown detection fixes (RH-BZ#746007)
by Stefan Bader 10 Apr '12

10 Apr '12

Resending this as there was no feedback overall and at least the first change fixes the following bug: https://bugzilla.redhat.com/show_bug.cgi?id=746007 --- The issue is easy to observe when using virt-manager as a frontend. Its main window seems to open connections for the up/down main display once and then only repeats DomainGetInfo on the same connection to check whether the instance is up or shut down. The initial open on the connection sets the domain id to -1 for inactive managed domains. And once running, the hypervisor driver returns the status. But after shutdown, the domain cannot be found by the hypervisor driver and the domain info is obtained by the xend driver. However, that expects the domain id to be -1 for inactive domains. This is not true (anymore). In testing I found that domain/status seems to be always present and better suited to check whether a domain is running or not (0 for shutdown and 2 for running, there are probably other values, too). Beside the change, which does fix this problem, the question is whether there should be a mechanism that resets domain->id to -1 when the instance is found to be shut down. And if that should happen, should it be done by any sub-driver or by the unified driver. The second problem I observed when I shut down an instance from the vnc console provided by virt-manager. As soon as the instance was down, the log produced internal errors about every second. Found this again, to be related to the unified driver approach. The call that causes the many error messages is GetVcpus. While the instance is up, the hypervisor driver returns those. But after shutdown, the domain cannot be found by that driver. The xend driver works, though. So what happens is that the hypervisor driver logs an internal error but the call succeeds because the xend driver returns successful. This generally opens the question whether a sub-driver should log any errors (at least not on error level, maybe debug)? The second patch only quiets down the one message I saw repeated that often. There are some other errors logged without any obvious problems interacting with the instances, but those are by far less often. -Stefan Please cc: me on replies as I am (yeah still) not subscribed to this m-l. Thanks.

2 10

[libvirt] [PATCH] tests: Fix libvirtdconftest in VPATH build
by Jiri Denemark 10 Apr '12

10 Apr '12

Without this, libvirtdconftest fails to build with "fatal error: daemon/libvirtd-config.h: No such file or directory" --- tests/Makefile.am | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/tests/Makefile.am b/tests/Makefile.am index 57358e9..c4d550f 100644 --- a/tests/Makefile.am +++ b/tests/Makefile.am @@ -6,6 +6,7 @@ SHELL = $(PREFERABLY_POSIX_SHELL) INCLUDES = \ + -I$(top_builddir) -I$(top_srcdir) \ -I$(top_builddir)/gnulib/lib -I$(top_srcdir)/gnulib/lib \ -I$(top_builddir)/include -I$(top_srcdir)/include \ -I$(top_builddir)/src -I$(top_srcdir)/src \ -- 1.7.8.5

2 1

[libvirt] libvirt starts KVM domain without -enable-kvm
by Reinier Schoof 10 Apr '12

10 Apr '12

Hey, we're running a VM-pool based on libvirt and QEMU/KVM. The host machines run debian 6 (squeeze). All VM's have a similar libvirt XML-definition. Since we've moved to the combination of Linux kernel 3.2.0-1 and libvirt 0.9.9 (installed from testing/wheezy repo's), we periodically notice poor performance of new(ly started) VM's and after inspection they turn out to be started without the qemu option -enable-kvm. In /var/log/libvirt/qemu/$VM-name.log a line like this is shown: 20529: error : qemuBuildCommandLine:3681 : unsupported configuration: the QEMU binary /usr/bin/kvm does not support kvm After a virsh destroy/start, using the same XML-definition, the VM is started WITH -enable-kvm. Looking at the libvirt source of 0.9.9, I can't imagine this being the result of a race condition or other strange circumstance: it simply parse 'qemu -help', which always give exactly the same output. I've tried reproducing the problem by continuously destroy, undefine, define and start the same VM and then grepping the kvm-commandline for the absence of -enable-kvm, without any result. The XML-definition includes <domain type='kvm'> and <emulator>/usr/bin/kvm</emulator> Does anyone recognize this behaviour of libvirt? Thanks, Reinier Schoof # virsh version Compiled against library: libvir 0.9.9 Using library: libvir 0.9.9 Using API: QEMU 0.9.9 Running hypervisor: QEMU 0.15.0 Linux 3.2.0-1-amd64 #1 SMP Fri Feb 17 05:17:36 UTC 2012 x86_64 GNU/Linux -- TransIP BV | https://www.transip.nl/

1 0

[libvirt] [PATCH] Fix comment about GNUTLS initialization/cleanup
by Daniel P. Berrange 10 Apr '12

10 Apr '12

From: "Daniel P. Berrange" <berrange(a)redhat.com> --- src/rpc/virnettlscontext.c | 10 +++++++--- 1 files changed, 7 insertions(+), 3 deletions(-) diff --git a/src/rpc/virnettlscontext.c b/src/rpc/virnettlscontext.c index 74a61e0..7440c7a 100644 --- a/src/rpc/virnettlscontext.c +++ b/src/rpc/virnettlscontext.c @@ -1423,9 +1423,13 @@ void virNetTLSSessionFree(virNetTLSSessionPtr sess) * virNetTLS* because it initializes * underlying GnuTLS library. According to * it's documentation, it's safe to be called - * many times, but is not thread safe. Each - * call SHOULD be later followed by - * virNetTLSContextDeinit. + * many times, but is not thread safe. + * + * There is no corresponding "Deinit" / "Cleanup" + * function because there is no safe way to call + * 'gnutls_global_deinit' from a multi-threaded + * library, where other libraries linked into the + * application may also be using gnutls. */ void virNetTLSInit(void) { -- 1.7.7.6

2 1

[libvirt] Memory leak in libvirt / gnutls
by Richard W.M. Jones 10 Apr '12

10 Apr '12

I'm trying to track down: https://bugzilla.redhat.com/show_bug.cgi?id=810613 I *think* this is a libvirt bug, but I can only reproduce it on one machine, not on another machine that has a virtually identical setup. Anyway, something is confusing me about libvirt. In this commit: ---------------------------------------------------------------------- commit 74c75671331d284e1f777f9692b72e9737520bf0 Author: Michal Privoznik <mprivozn(a)redhat.com> Date: Thu Aug 18 10:44:08 2011 +0200 daemon: initialize GnuTLS When spice_tls is set but listen_tls is not, we don't initialize GnuTLS library. So any later gnutls call (e.g. during migration, where we initialize a certificate) will access uninitialized GnuTLS internal structs and throws an error. Although, we might now initialize GnuTLS twice, it is safe according to the documentation: This function can be called many times, but will only do something the first time. This patch creates 2 functions: virNetTLSInit and virNetTLSDeinit with respect to written above. ---------------------------------------------------------------------- ... a pair of functions called virNetTLSInit & virNetTLSDeinit are introduced, which would appear to do the right thing, calling first gnutls_global_init, then gnutls_global_deinit. This looks correct to me. However the above commit is later amended by this commit: ---------------------------------------------------------------------- commit eaddec976ef06457fee4a4ce86b8c7ee906183b7 Author: Michal Privoznik <mprivozn(a)redhat.com> Date: Wed Aug 24 16:16:45 2011 +0200 daemon: Move TLS initialization to virInitialize My previous patch 74c75671331d284e1f777f9692b72e9737520bf0 introduced a regression by removing TLS initialization from client. ---------------------------------------------------------------------- which removes virNetTLSDeinit. This appears to be a mistake, or at least I can't see the logical reason for it, and according to the gnutls docs, it would introduce a memory leak looking exactly like the one I am chasing down. The current code mentions virNetTLSDeinit only in a comment, and never calls gnutls_global_deinit. Any ideas? Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones virt-top is 'top' for virtual machines. Tiny program with many powerful monitoring features, net stats, disk stats, logging, etc. http://et.redhat.com/~rjones/virt-top

3 3

[libvirt] [PATCH] Fix compilation error on 32bit
by Stefan Berger 10 Apr '12

10 Apr '12

Below code failed to compile on a 32 bit machine with error typewrappers.c: In function 'libvirt_intUnwrap': typewrappers.c:135:5: error: logical 'and' of mutually exclusive tests is always false [-Werror=logical-op] cc1: all warnings being treated as errors The patch fixes this error. Stefan --- python/typewrappers.c | 4 ++++ 1 file changed, 4 insertions(+) Index: libvirt-acl/python/typewrappers.c =================================================================== --- libvirt-acl.orig/python/typewrappers.c +++ libvirt-acl/python/typewrappers.c @@ -132,6 +132,7 @@ libvirt_intUnwrap(PyObject *obj, int *va if ((long_val == -1) && PyErr_Occurred()) return -1; +#if LONG_MAX != INT_MAX if (long_val >= INT_MIN && long_val <= INT_MAX) { *val = long_val; } else { @@ -139,6 +140,9 @@ libvirt_intUnwrap(PyObject *obj, int *va "Python int too large to convert to C int"); return -1; } +#else + *val = long_val; +#endif return 0; }

2 2

[libvirt] [PATCH 0/4] Replace daemon-conf test case
by Daniel P. Berrange 10 Apr '12

10 Apr '12

The daemon-conf test case is terminally broken by design, despite numerous patches over the years, due to its need to test functionality indirectly by running libvirtd. Replace it with a test case which directly unit tests the code in question.

2 9

[libvirt] [test-API PATCH 1/3] test-API: Get rid of utils/Python
by Osier Yang 10 Apr '12

10 Apr '12

test-API tends to only support testing via libvirt Python binding now, and moving it to utils makes sense. --- utils/{Python => }/XMLParser.py | 0 utils/{Python => }/check.py | 0 utils/{Python => }/env_parser.py | 0 utils/{Python => }/format.py | 0 utils/{Python => }/log.py | 0 utils/{Python => }/utils.py | 0 utils/{Python => }/xmlbuilder.py | 0 utils/{Python => }/xmlgenerator.py | 0 8 files changed, 0 insertions(+), 0 deletions(-) delete mode 100644 utils/Python/__init__.py rename utils/{Python => }/XMLParser.py (100%) rename utils/{Python => }/check.py (100%) rename utils/{Python => }/env_parser.py (100%) rename utils/{Python => }/format.py (100%) rename utils/{Python => }/log.py (100%) rename utils/{Python => }/utils.py (100%) rename utils/{Python => }/xmlbuilder.py (100%) rename utils/{Python => }/xmlgenerator.py (100%) diff --git a/utils/Python/__init__.py b/utils/Python/__init__.py deleted file mode 100644 index e69de29..0000000 diff --git a/utils/Python/XMLParser.py b/utils/XMLParser.py similarity index 100% rename from utils/Python/XMLParser.py rename to utils/XMLParser.py diff --git a/utils/Python/check.py b/utils/check.py similarity index 100% rename from utils/Python/check.py rename to utils/check.py diff --git a/utils/Python/env_parser.py b/utils/env_parser.py similarity index 100% rename from utils/Python/env_parser.py rename to utils/env_parser.py diff --git a/utils/Python/format.py b/utils/format.py similarity index 100% rename from utils/Python/format.py rename to utils/format.py diff --git a/utils/Python/log.py b/utils/log.py similarity index 100% rename from utils/Python/log.py rename to utils/log.py diff --git a/utils/Python/utils.py b/utils/utils.py similarity index 100% rename from utils/Python/utils.py rename to utils/utils.py diff --git a/utils/Python/xmlbuilder.py b/utils/xmlbuilder.py similarity index 100% rename from utils/Python/xmlbuilder.py rename to utils/xmlbuilder.py diff --git a/utils/Python/xmlgenerator.py b/utils/xmlgenerator.py similarity index 100% rename from utils/Python/xmlgenerator.py rename to utils/xmlgenerator.py -- 1.7.7.5

2 5

[libvirt] help for libvirt0.9.8 compile
by hero 10 Apr '12

10 Apr '12

my redhat6.1 has a libvirt0.8.7, and I got a libvirt0.9.8.tar that I want to update the old one. when I use './configure --prefix=/usr --libdir=/usr/lib64 --localstatedir=/var --sysconfdir=/etc' it siad 'configure: error: You must install device-mapper-devel/libdevmapper >= 1.0.0 to compile libvirt' please tell me how to do if you know. thanks. best wishes! SunYongGang

3 2

Re: [libvirt] Fwd: Look what i found...
by dasari dakshina 10 Apr '12

10 Apr '12

<p>How to start an Online Business<br><a href="http://expertizaregion55.ru/breakingnews/48IanStewart/">http://expertizaregion55.ru/breakingnews/48IanStewart/</a></p>

1 0

[libvirt] [PATCHv3 00/16] live block migration via virDomainBlockCopy
by Eric Blake 09 Apr '12

09 Apr '12

v1 was here: https://www.redhat.com/archives/libvir-list/2012-April/msg00068.html v2 was here: https://www.redhat.com/archives/libvir-list/2012-April/msg00222.html changes from v2: added patch 12/16, addressed some review comments and made minor changes from more of my own testing CAVEAT: Paolo and I had an IRC conversation, where we decided that 'drive-mirror' might do better to take a mandatory 'full':'bool' (or an even more powerful optional '*base':'str') argument for determining how much to stream (full vs. shallow, or even the backing file of the copy as matching the base argument of 'block_stream') rather than overloading the 'mode':'no-backing-file' as the only way to get a full pull. If that change goes in qemu, then we will need a v4, to tweak how the 'drive-mirror' monitor command is called, and take advantage of the additional flexibility it offers. changes from v1: Paolo has updated the qemu side of things, and built a scratch image for RHEL that I was able to test with for the new semantics. I was actually able to successfully run two different block copy jobs, one canceled and one pivoted, with SELinux disabled and this patch series. Patch 13/16 is for reference only when working with Paolo's build; it will not go upstream. Patches 14-16 are optional; the extra flexibility might be nice, but I haven't yet played with those three enough to know if Paolo's build behaves like I was expecting. Adam Litke (2): blockjob: add API for async virDomainBlockJobAbort blockjob: wire up qemu async virDomainBlockJobAbort Eric Blake (14): blockjob: allow for fast-finishing job blockjob: add new API flags blockjob: add 'blockcopy' to virsh blockjob: enhance xml to track mirrors across libvirtd restart blockjob: react to active block copy blockjob: expose qemu commands for mirrored storage migration blockjob: return appropriate event and info blockjob: support pivot operation on cancel blockjob: implement block copy for qemu blockjob: allow for existing files blockjob: accommodate RHEL backport names blockjob: add virDomainBlockCopy blockjob: enhance virsh 'blockcopy' blockjob: wire up qemu and RPC for block copy docs/apibuild.py | 1 + docs/formatdomain.html.in | 11 ++ docs/schemas/domaincommon.rng | 19 ++- include/libvirt/libvirt.h.in | 50 +++++- include/libvirt/virterror.h | 1 + src/conf/domain_conf.c | 54 ++++++ src/conf/domain_conf.h | 6 + src/driver.h | 5 + src/libvirt.c | 220 +++++++++++++++++++++++- src/libvirt_private.syms | 1 + src/libvirt_public.syms | 5 + src/qemu/qemu_capabilities.c | 3 + src/qemu/qemu_capabilities.h | 2 + src/qemu/qemu_driver.c | 379 +++++++++++++++++++++++++++++++++++++++-- src/qemu/qemu_hotplug.c | 7 + src/qemu/qemu_monitor.c | 50 ++++++ src/qemu/qemu_monitor.h | 26 +++- src/qemu/qemu_monitor_json.c | 166 +++++++++++++++--- src/qemu/qemu_monitor_json.h | 19 ++- src/remote/remote_driver.c | 1 + src/remote/remote_protocol.x | 12 ++- src/remote_protocol-structs | 9 + src/rpc/gendispatch.pl | 1 + src/util/virterror.c | 6 + tools/virsh.c | 140 ++++++++++++--- tools/virsh.pod | 40 ++++- 26 files changed, 1143 insertions(+), 91 deletions(-) -- 1.7.7.6

2 20