- Devel - libvirt List Archives

[libvirt] Reminder: KVM Forum 2012 Call For Participation
by KVM Forum 2012 Program Committee 27 Aug '12

27 Aug '12

Just a reminder, the CFP ends this Friday. -- ================================================================= KVM Forum 2012: Call For Participation November 7-9, 2012 - Hotel Fira Palace - Barcelona, Spain (All submissions must be received before midnight Aug 31st, 2012) ================================================================= KVM is an industry leading open source hypervisor that provides an ideal platform for datacenter virtualization, virtual desktop infrastructure, and cloud computing. Once again, it's time to bring together the community of developers and users that define the KVM ecosystem for our annual technical conference. We will discuss the current state of affairs and plan for the future of KVM, its surrounding infrastructure, and management tools. We are also excited to announce the oVirt Workshop will run in parallel with the KVM Forum, bringing in a community focused on enterprise datacenter virtualization management built on KVM. For topics which overlap we will have shared sessions. So mark your calendar and join us in advancing KVM. http://events.linuxfoundation.org/events/kvm-forum/ Once again we are colocated with The Linux Foundation's LinuxCon, Based on feedback from last year, this time it's LinuxCon Europe! KVM Forum attendees will be able to attend oVirt Workshop sessions and are eligible to attend LinuxCon Europe for a discounted rate. http://events.linuxfoundation.org/events/kvm-forum/register We invite you to lead part of the discussion by submitting a speaking proposal for KVM Forum 2012. http://events.linuxfoundation.org/cfp Suggested topics: KVM - Scaling and performance - Nested virtualization - I/O improvements - PCI device assignment - Driver domains - Time keeping - Resource management (cpu, memory, i/o) - Memory management (page sharing, swapping, huge pages, etc) - VEPA, VN-Link, vswitch - Security - Architecture ports QEMU - Device model improvements - New devices and chipsets - Scaling and performance - Desktop virtualization - Spice - Increasing robustness and hardening - Security model - Management interfaces - QMP protocol and implementation - Image formats - Firmware (SeaBIOS, OVMF, UEFI, etc) - Live migration - Live snapshots and merging - Fault tolerance, high availability, continuous backup - Real-time guest support Virtio - Speeding up existing devices - Alternatives - Virtio on non-Linux or non-virtualized Management infrastructure - oVirt (shared track w/ oVirt Workshop) - Libvirt - KVM autotest - OpenStack - Network virtualization management - Enterprise storage management Cloud computing - Scalable storage - Virtual networking - Security - Provisioning SUBMISSION REQUIREMENTS Abstracts due: Aug 31st, 2012 Notification: Sep 14th, 2012 Please submit a short abstract (~150 words) describing your presentation proposal. In your submission please note how long your talk will take. Slots vary in length up to 45 minutes. Also include in your proposal the proposal type -- one of: - technical talk - end-user talk - birds of a feather (BOF) session Submit your proposal here: http://events.linuxfoundation.org/cfp You will receive a notification whether or not your presentation proposal was accepted by Sep 14th. END-USER COLLABORATION One of the big challenges as developers is to know what, where and how people actually use our software. We will reserve a few slots for end users talking about their deployment challenges and achievements. If you are using KVM in production you are encouraged submit a speaking proposal. Simply mark it as an end-user collaboration proposal. As an end user, this is a unique opportunity to get your input to developers. BOF SESSION We will reserve some slots in the evening after the main conference tracks, for birds of a feather (BOF) sessions. These sessions will be less formal than presentation tracks and targetted for people who would like to discuss specific issues with other developers and/or users. If you are interested in getting developers and/or uses together to discuss a specific problem, please submit a BOF proposal. LIGHTNING TALKS In addition to submitted talks we will also have some room for lightning talks. These are short (5 minute) discussions to highlight new work or ideas that aren't complete enough to warrant a full presentation slot. Lightning talk submissions and scheduling will be handled on-site at KVM Forum. HOTEL / TRAVEL The KVM Forum 2012 will be held in Barcelona, Spain at the Hotel Fira Palace. http://events.linuxfoundation.org/events/kvm-forum/hotel Thank you for your interest in KVM. We're looking forward to your submissions and seeing you at the KVM Forum 2012 in November! Thanks, your KVM Forum 2012 Program Commitee Please contact us with any questions or comments. KVM-Forum-2012-PC(a)redhat.com

1 0

[libvirt] [PATCH] qemu: Fix define logic
by Michal Privoznik 27 Aug '12

27 Aug '12

With current flow in qemudDomainDefine we might loose data when updating an existing domain. We parse given XML and overwrite the configuration. Then we try to save the new config. However, this step may fail and we don't perform any roll back. In fact, we remove the domain from the list of domains held up by qemu driver. This is okay as long as the domain was brand new one. --- The easiest steps to reproduce: https://bugzilla.redhat.com/show_bug.cgi?id=851963 src/qemu/qemu_driver.c | 43 ++++++++++++++++++++++++++++++++++++------- 1 files changed, 36 insertions(+), 7 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 3fcca0e..e760ed9 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -5538,6 +5538,7 @@ out: static virDomainPtr qemudDomainDefine(virConnectPtr conn, const char *xml) { struct qemud_driver *driver = conn->privateData; virDomainDefPtr def; + virDomainDefPtr def_backup = NULL; virDomainObjPtr vm = NULL; virDomainPtr dom = NULL; virDomainEventPtr event = NULL; @@ -5561,20 +5562,48 @@ static virDomainPtr qemudDomainDefine(virConnectPtr conn, const char *xml) { if (qemuDomainAssignAddresses(def, NULL, NULL) < 0) goto cleanup; - if (!(vm = virDomainAssignDef(driver->caps, - &driver->domains, - def, false))) { - goto cleanup; + /* We need to differentiate two cases: + * a) updating an existing domain - must preserve previous definition + * so we can roll back if something fails + * b) defining a brand new domain - virDomainAssignDef is just sufficient + */ + if ((vm = virDomainFindByUUID(&driver->domains, def->uuid))) { + if (virDomainObjIsActive(vm)) { + def_backup = vm->newDef; + vm->newDef = def; + } else { + def_backup = vm->def; + vm->def = def; + } + } else { + if (!(vm = virDomainAssignDef(driver->caps, + &driver->domains, + def, false))) { + goto cleanup; + } } def = NULL; vm->persistent = 1; if (virDomainSaveConfig(driver->configDir, vm->newDef ? vm->newDef : vm->def) < 0) { - VIR_INFO("Defining domain '%s'", vm->def->name); - qemuDomainRemoveInactive(driver, vm); - vm = NULL; + if (def_backup) { + /* There is backup so this VM was defined before. + * Just restore the backup. */ + VIR_INFO("Restoring domain '%s' definition", vm->def->name); + if (virDomainObjIsActive(vm)) + vm->newDef = def_backup; + else + vm->def = def_backup; + } else { + /* Brand new domain. Remove it */ + VIR_INFO("Deleting domain '%s'", vm->def->name); + qemuDomainRemoveInactive(driver, vm); + vm = NULL; + } goto cleanup; + } else { + virDomainDefFree(def_backup); } event = virDomainEventNewFromObj(vm, -- 1.7.8.6

2 2

[libvirt] [PATCH 1/2] Introduce new VIR_ERR_AGENT_UNRESPONSIVE error code
by Michal Privoznik 27 Aug '12

27 Aug '12

Currently, when guest agent is configure but not responsive (e.g. due to appropriate service not running in the guest) we return VIR_ERR_INTERNAL_ERROR or VIR_ERR_ARGUMENT_UNSUPPORTED. Both are wrong. Therefore we need to introduce new error code to reflect this case. --- include/libvirt/virterror.h | 2 ++ src/qemu/qemu_agent.c | 2 +- src/qemu/qemu_driver.c | 24 ++++++++++++++---------- src/util/virterror.c | 7 +++++++ 4 files changed, 24 insertions(+), 11 deletions(-) diff --git a/include/libvirt/virterror.h b/include/libvirt/virterror.h index 69c64aa..5140c38 100644 --- a/include/libvirt/virterror.h +++ b/include/libvirt/virterror.h @@ -283,6 +283,8 @@ typedef enum { VIR_ERR_OPERATION_UNSUPPORTED = 84, /* The requested operation is not supported */ VIR_ERR_SSH = 85, /* error in ssh transport driver */ + VIR_ERR_AGENT_UNRESPONSIVE = 86, /* guest agent is unresponsive, + not running or not usable */ } virErrorNumber; /** diff --git a/src/qemu/qemu_agent.c b/src/qemu/qemu_agent.c index c658bf8..ba29783 100644 --- a/src/qemu/qemu_agent.c +++ b/src/qemu/qemu_agent.c @@ -880,7 +880,7 @@ static int qemuAgentSend(qemuAgentPtr mon, if ((timeout && virCondWaitUntil(&mon->notify, &mon->lock, then) < 0) || (!timeout && virCondWait(&mon->notify, &mon->lock) < 0)) { if (errno == ETIMEDOUT) { - virReportError(VIR_ERR_INTERNAL_ERROR, "%s", + virReportError(VIR_ERR_AGENT_UNRESPONSIVE, "%s", _("Guest agent not available for now")); ret = -2; } else { diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 3fcca0e..f64d9ec 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -1734,8 +1734,9 @@ static int qemuDomainShutdownFlags(virDomainPtr dom, unsigned int flags) { if (useAgent) { if (priv->agentError) { - virReportError(VIR_ERR_INTERNAL_ERROR, "%s", - _("QEMU guest agent is not available due to an error")); + virReportError(VIR_ERR_AGENT_UNRESPONSIVE, "%s", + _("QEMU guest agent is not " + "available due to an error")); goto cleanup; } if (!priv->agent) { @@ -1815,8 +1816,9 @@ qemuDomainReboot(virDomainPtr dom, unsigned int flags) if (useAgent) { if (priv->agentError) { - virReportError(VIR_ERR_INTERNAL_ERROR, "%s", - _("QEMU guest agent is not available due to an error")); + virReportError(VIR_ERR_AGENT_UNRESPONSIVE, "%s", + _("QEMU guest agent is not " + "available due to an error")); goto cleanup; } if (!priv->agent) { @@ -10391,7 +10393,7 @@ qemuDomainSnapshotFSFreeze(struct qemud_driver *driver, int freezed; if (priv->agentError) { - virReportError(VIR_ERR_INTERNAL_ERROR, "%s", + virReportError(VIR_ERR_AGENT_UNRESPONSIVE, "%s", _("QEMU guest agent is not " "available due to an error")); return -1; @@ -10419,7 +10421,7 @@ qemuDomainSnapshotFSThaw(struct qemud_driver *driver, if (priv->agentError) { if (report) - virReportError(VIR_ERR_INTERNAL_ERROR, "%s", + virReportError(VIR_ERR_AGENT_UNRESPONSIVE, "%s", _("QEMU guest agent is not " "available due to an error")); return -1; @@ -13708,8 +13710,9 @@ qemuDomainPMSuspendForDuration(virDomainPtr dom, } if (priv->agentError) { - virReportError(VIR_ERR_INTERNAL_ERROR, "%s", - _("QEMU guest agent is not available due to an error")); + virReportError(VIR_ERR_AGENT_UNRESPONSIVE, "%s", + _("QEMU guest agent is not " + "available due to an error")); goto cleanup; } @@ -13849,8 +13852,9 @@ qemuDomainAgentCommand(virDomainPtr domain, } if (priv->agentError) { - virReportError(VIR_ERR_INTERNAL_ERROR, "%s", - _("QEMU guest agent is not available due to an error")); + virReportError(VIR_ERR_AGENT_UNRESPONSIVE, "%s", + _("QEMU guest agent is not " + "available due to an error")); goto cleanup; } diff --git a/src/util/virterror.c b/src/util/virterror.c index 3ee2ae0..7caa69e 100644 --- a/src/util/virterror.c +++ b/src/util/virterror.c @@ -1199,6 +1199,13 @@ virErrorMsg(virErrorNumber error, const char *info) errmsg = _("SSH transport error"); else errmsg = _("SSH transport error: %s"); + break; + case VIR_ERR_AGENT_UNRESPONSIVE: + if (info == NULL) + errmsg = _("Guest agent is not responding"); + else + errmsg = _("Guest agent is not responding: %s"); + break; } return errmsg; } -- 1.7.8.6

2 4

[libvirt] [PATCH libvirt-glib] Don't hold events lock when dispatching free callbacks
by Daniel P. Berrange 27 Aug '12

27 Aug '12

From: "Daniel P. Berrange" <berrange(a)redhat.com> The _event_timeout_remove and _event_handle_remove methods were holding onto the global lock when invoking the free callback. This is a violation of the libvirt events API contract which requires free callbacks to be invoked in a re-entrant safe context. --- libvirt-glib/libvirt-glib-event.c | 8 ++------ 1 file changed, 2 insertions(+), 6 deletions(-) diff --git a/libvirt-glib/libvirt-glib-event.c b/libvirt-glib/libvirt-glib-event.c index 657c1bf..2a9ee23 100644 --- a/libvirt-glib/libvirt-glib-event.c +++ b/libvirt-glib/libvirt-glib-event.c @@ -256,13 +256,11 @@ _event_handle_remove(gpointer data) { struct gvir_event_handle *h = data; - g_mutex_lock(eventlock); - if (h->ff) (h->ff)(h->opaque); + g_mutex_lock(eventlock); g_ptr_array_remove_fast(handles, h); - g_mutex_unlock(eventlock); return FALSE; @@ -414,13 +412,11 @@ _event_timeout_remove(gpointer data) { struct gvir_event_timeout *t = data; - g_mutex_lock(eventlock); - if (t->ff) (t->ff)(t->opaque); + g_mutex_lock(eventlock); g_ptr_array_remove_fast(timeouts, t); - g_mutex_unlock(eventlock); return FALSE; -- 1.7.11.2

2 1

[libvirt] [PATCH] conf: prevent NULL pointer access in virSecurityLabelDefsParseXML
by Ján Tomko 27 Aug '12

27 Aug '12

When checking for seclabels without security models, def->nseclabels is already set to n. In the case of an error def->seclabels is freed but nseclabels is left untouched. This leads to a segmentation fault when def is freed in virDomainDefParseXML. --- src/conf/domain_conf.c | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 82ccf4d..c02d6f8 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -3179,6 +3179,7 @@ error: virSecurityLabelDefFree(def->seclabels[i - 1]); } VIR_FREE(def->seclabels); + def->nseclabels = 0; VIR_FREE(list); return -1; } -- 1.7.8.6

2 1

[libvirt] [PATCH] qemu: Switch to unified func name
by Michal Privoznik 27 Aug '12

27 Aug '12

With the latest patches libvirt supports qemu agent monitor passthrough. However, function in qemu driver is called qemuDrvDomainAgentCommand. s/Drv// as used in all other names. --- src/qemu/qemu_driver.c | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 64c407d..3fcca0e 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -13815,7 +13815,7 @@ qemuListAllDomains(virConnectPtr conn, } static char * -qemuDrvDomainAgentCommand(virDomainPtr domain, +qemuDomainAgentCommand(virDomainPtr domain, const char *cmd, int timeout, unsigned int flags) @@ -14028,7 +14028,7 @@ static virDriver qemuDriver = { .domainSnapshotDelete = qemuDomainSnapshotDelete, /* 0.8.0 */ .qemuDomainMonitorCommand = qemuDomainMonitorCommand, /* 0.8.3 */ .qemuDomainAttach = qemuDomainAttach, /* 0.9.4 */ - .qemuDomainArbitraryAgentCommand = qemuDrvDomainAgentCommand, /* 0.10.0 */ + .qemuDomainArbitraryAgentCommand = qemuDomainAgentCommand, /* 0.10.0 */ .domainOpenConsole = qemuDomainOpenConsole, /* 0.8.6 */ .domainOpenGraphics = qemuDomainOpenGraphics, /* 0.9.7 */ .domainInjectNMI = qemuDomainInjectNMI, /* 0.9.2 */ -- 1.7.8.6

2 2

[libvirt] [PATCH] rpc: fix segmentation fault caused by null client-sock
by Guannan Ren 27 Aug '12

27 Aug '12

BZ:https://bugzilla.redhat.com/show_bug.cgi?id=851423 The client-sock could have been set to NULL by eventloop thread after async event fired. --- src/rpc/virnetclient.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/src/rpc/virnetclient.c b/src/rpc/virnetclient.c index 4ecc703..43a9814 100644 --- a/src/rpc/virnetclient.c +++ b/src/rpc/virnetclient.c @@ -644,7 +644,8 @@ virNetClientMarkClose(virNetClientPtr client, int reason) { VIR_DEBUG("client=%p, reason=%d", client, reason); - virNetSocketRemoveIOCallback(client->sock); + if (client->sock) + virNetSocketRemoveIOCallback(client->sock); client->wantClose = true; client->closeReason = reason; } -- 1.7.11.4

2 2

[libvirt] [PATCH] qemu: fix regression with pinning
by Martin Kletzander 27 Aug '12

27 Aug '12

Commit 4b03d59167f4a4c6ec57def315a61d977466e75b changed the pinning behavior in a way that makes some machine non-startable. The comment mentioning that we cannot control each vcpu when there is no VCPU<->PID mapping available is true, however, this isn't necessarily an error, because this can be caused by old QEMU without support for "query-cpus" command as well as a software emulated machines that don't create more than one process. --- src/qemu/qemu_cgroup.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/src/qemu/qemu_cgroup.c b/src/qemu/qemu_cgroup.c index 2237d11..b45bb49 100644 --- a/src/qemu/qemu_cgroup.c +++ b/src/qemu/qemu_cgroup.c @@ -567,9 +567,9 @@ int qemuSetupCgroupForVcpu(struct qemud_driver *driver, virDomainObjPtr vm) /* If we don't know VCPU<->PID mapping or all vcpu runs in the same * thread, we cannot control each vcpu. */ - virReportError(VIR_ERR_INTERNAL_ERROR, "%s", - _("Unable to get vcpus' pids.")); - goto cleanup; + VIR_DEBUG("%s", _("Unable to get vcpus' pids.")); + virCgroupFree(&cgroup); + return 0; } for (i = 0; i < priv->nvcpupids; i++) { -- 1.7.12

2 2

[libvirt] [PATCH] qemu: fix regression with spice tls port allocation
by Martin Kletzander 27 Aug '12

27 Aug '12

In my quest for reusing variables I failed to edit one variable when fixing details between two patch versions. That results in a failure to start qemu with autoport and spice tls, because qemu is trying to bind two sockets to the same port. --- src/qemu/qemu_process.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index abe6e74..7f85aea 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -3500,7 +3500,7 @@ int qemuProcessStart(virConnectPtr conn, goto cleanup; } - vm->def->graphics[0]->data.spice.tlsPort = port; + vm->def->graphics[0]->data.spice.tlsPort = tlsPort; } } -- 1.7.12

2 2

[libvirt] [PATCH] virsh: fix missing return value
by Alex Jia 27 Aug '12

27 Aug '12

Although virsh command raises a correct error information, the command status returns 0(true), this patch is used for fixing this issue. Signed-off-by: Alex Jia <ajia(a)redhat.com> --- tools/virsh-host.c | 3 +++ 1 files changed, 3 insertions(+), 0 deletions(-) diff --git a/tools/virsh-host.c b/tools/virsh-host.c index da60895..2f3b413 100644 --- a/tools/virsh-host.c +++ b/tools/virsh-host.c @@ -701,11 +701,13 @@ cmdQemuAgentCommand(vshControl *ctl, const vshCmd *cmd) judge = vshCommandOptInt(cmd, "timeout", &timeout); if (judge < 0) { vshError(ctl, "%s", _("timeout number has to be a number")); + goto cleanup; } else if (judge > 0) { judge = 1; } if (judge && timeout < 1) { vshError(ctl, "%s", _("timeout must be positive")); + goto cleanup; } if (vshCommandOptBool(cmd, "async")) { @@ -719,6 +721,7 @@ cmdQemuAgentCommand(vshControl *ctl, const vshCmd *cmd) if (judge > 1) { vshError(ctl, "%s", _("timeout, async and block options are exclusive")); + goto cleanup; } result = virDomainQemuAgentCommand(dom, guest_agent_cmd, timeout, flags); -- 1.7.1

2 2

[libvirt] [PATCH] specfile: require libnl3 for Fedora >= 18 and RHEL >= 7
by Laine Stump 27 Aug '12

27 Aug '12

Everything is ready in both netcf and libvirt to switch over to libnl3 in future releases of both Fedora and RHEL. This needs to be done more or less simultaneously in both packages, though, because you can't mix libnl1.1 and libnl3 in the same process (e.g. libvirtd using libnl-3.so and libnetcf.so, while libnetcf.so uses libnl.so) This patch does two things when fedora >= 18 || rhel >= 7): 1) requires libnl3-devel 2) requires netcf-devel-0.2.2 or greater (the idea is that a similar patch is going into netcf's specfile, so that when a build of netcf is done on F18 or later (or RHEL7 or later) netcf will be guaranteed to be built with libnl3 rather than libnl-1.1) --- libvirt.spec.in | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/libvirt.spec.in b/libvirt.spec.in index 18a7fb8..4b4d1d0 100644 --- a/libvirt.spec.in +++ b/libvirt.spec.in @@ -409,8 +409,12 @@ BuildRequires: sanlock-devel >= 1.8 BuildRequires: libpcap-devel %endif %if %{with_libnl} +%if 0%{?fedora} >= 18 || 0%{?rhel} >= 7 +BuildRequires: libnl3-devel +%else BuildRequires: libnl-devel %endif +%endif %if %{with_avahi} BuildRequires: avahi-devel %endif @@ -489,13 +493,18 @@ BuildRequires: libcap-ng-devel >= 0.5.0 %if %{with_phyp} BuildRequires: libssh2-devel %endif + %if %{with_netcf} +%if 0%{?fedora} >= 18 || 0%{?rhel} >= 7 +BuildRequires: netcf-devel >= 0.2.2 +%else %if 0%{?fedora} >= 16 || 0%{?rhel} >= 6 BuildRequires: netcf-devel >= 0.1.8 %else BuildRequires: netcf-devel >= 0.1.4 %endif %endif +%endif %if %{with_esx} %if 0%{?fedora} >= 9 || 0%{?rhel} >= 6 BuildRequires: libcurl-devel -- 1.7.11.4

4 7

[libvirt] [PATCH v2] esx: Support VLAN tags in virtual network port groups
by Matthias Bolte 26 Aug '12

26 Aug '12

--- v2: Use network level VLAN config if there is no portgroup specific VLAN config given. src/esx/esx_network_driver.c | 65 ++++++++++++++++++++++++++++++++++++++--- 1 files changed, 60 insertions(+), 5 deletions(-) diff --git a/src/esx/esx_network_driver.c b/src/esx/esx_network_driver.c index 09d46d3..7b529ee 100644 --- a/src/esx/esx_network_driver.c +++ b/src/esx/esx_network_driver.c @@ -489,7 +489,40 @@ esxNetworkDefineXML(virConnectPtr conn, const char *xml) goto cleanup; } - hostPortGroupSpec->vlanId->value = 0; + if (def->portGroups[i].vlan.trunk) { + /* FIXME: Change this once tag-less trunk-mode is supported */ + if (def->portGroups[i].vlan.nTags != 1 || + *def->portGroups[i].vlan.tag != 4095) { + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", + _("VLAN tag has to be 4095 in trunk mode")); + goto cleanup; + } + + hostPortGroupSpec->vlanId->value = 4095; + } else if (def->portGroups[i].vlan.nTags == 1) { + hostPortGroupSpec->vlanId->value = *def->portGroups[i].vlan.tag; + } else if (def->portGroups[i].vlan.nTags > 1) { + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", + _("Can apply one VLAN tag per port group only")); + goto cleanup; + } else if (def->vlan.trunk) { + /* FIXME: Change this once tag-less trunk-mode is supported */ + if (def->vlan.nTags != 1 || *def->vlan.tag != 4095) { + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", + _("VLAN tag has to be 4095 in trunk mode")); + goto cleanup; + } + + hostPortGroupSpec->vlanId->value = 4095; + } else if (def->vlan.nTags == 1) { + hostPortGroupSpec->vlanId->value = *def->vlan.tag; + } else if (def->vlan.nTags > 1) { + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", + _("Can apply one VLAN tag per port group only")); + goto cleanup; + } else { + hostPortGroupSpec->vlanId->value = 0; + } if (def->portGroups[i].bandwidth != NULL) { if (esxBandwidthToShapingPolicy @@ -519,6 +552,8 @@ esxNetworkDefineXML(virConnectPtr conn, const char *xml) network = virGetNetwork(conn, hostVirtualSwitch->name, md5); cleanup: + /* FIXME: need to remove virtual switch if adding port groups failed */ + virNetworkDefFree(def); esxVI_HostVirtualSwitch_Free(&hostVirtualSwitch); esxVI_HostPortGroup_Free(&hostPortGroupList); @@ -695,6 +730,7 @@ esxNetworkGetXMLDesc(virNetworkPtr network_, unsigned int flags) esxVI_String *hostPortGroupKey = NULL; esxVI_String *networkName = NULL; virNetworkDefPtr def; + virPortGroupDefPtr portGroup; if (esxVI_EnsureSession(priv->primary) < 0) { return NULL; @@ -824,9 +860,12 @@ esxNetworkGetXMLDesc(virNetworkPtr network_, unsigned int flags) for (networkName = networkNameList; networkName != NULL; networkName = networkName->_next) { if (STREQ(networkName->value, hostPortGroup->spec->name)) { - def->portGroups[def->nPortGroups].name = strdup(networkName->value); + portGroup = &def->portGroups[def->nPortGroups]; + ++def->nPortGroups; - if (def->portGroups[def->nPortGroups].name == NULL) { + portGroup->name = strdup(networkName->value); + + if (portGroup->name == NULL) { virReportOOMError(); goto cleanup; } @@ -834,13 +873,29 @@ esxNetworkGetXMLDesc(virNetworkPtr network_, unsigned int flags) if (hostPortGroup->spec->policy != NULL) { if (esxShapingPolicyToBandwidth (hostPortGroup->spec->policy->shapingPolicy, - &def->portGroups[def->nPortGroups].bandwidth) < 0) { + &portGroup->bandwidth) < 0) { ++def->nPortGroups; goto cleanup; } } - ++def->nPortGroups; + if (hostPortGroup->spec->vlanId->value > 0) { + if (hostPortGroup->spec->vlanId->value == 4095) { + portGroup->vlan.trunk = true; + } + + /* FIXME: Remove this once tag-less trunk-mode is supported */ + portGroup->vlan.nTags = 1; + + if (VIR_ALLOC_N(portGroup->vlan.tag, 1) < 0) { + virReportOOMError(); + goto cleanup; + } + + *portGroup->vlan.tag = + hostPortGroup->spec->vlanId->value; + } + break; } } -- 1.7.4.1

2 1

[libvirt] [PATCH] esx: Support VLAN tags in virtual network port groups
by Matthias Bolte 26 Aug '12

26 Aug '12

--- src/esx/esx_network_driver.c | 38 +++++++++++++++++++++++++++++++++----- 1 files changed, 33 insertions(+), 5 deletions(-) diff --git a/src/esx/esx_network_driver.c b/src/esx/esx_network_driver.c index 09d46d3..2f5f1ab 100644 --- a/src/esx/esx_network_driver.c +++ b/src/esx/esx_network_driver.c @@ -489,7 +489,16 @@ esxNetworkDefineXML(virConnectPtr conn, const char *xml) goto cleanup; } - hostPortGroupSpec->vlanId->value = 0; + if (def->portGroups[i].vlan.trunk) { + hostPortGroupSpec->vlanId->value = 4095; + } else if (def->portGroups[i].vlan.nTags == 1) { + hostPortGroupSpec->vlanId->value = *def->portGroups[i].vlan.tag; + } else if (def->portGroups[i].vlan.nTags > 1) { + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", + _("Can apply one VLAN tag per port group only")); + } else { + hostPortGroupSpec->vlanId->value = 0; + } if (def->portGroups[i].bandwidth != NULL) { if (esxBandwidthToShapingPolicy @@ -695,6 +704,7 @@ esxNetworkGetXMLDesc(virNetworkPtr network_, unsigned int flags) esxVI_String *hostPortGroupKey = NULL; esxVI_String *networkName = NULL; virNetworkDefPtr def; + virPortGroupDefPtr portGroup = NULL; if (esxVI_EnsureSession(priv->primary) < 0) { return NULL; @@ -824,9 +834,12 @@ esxNetworkGetXMLDesc(virNetworkPtr network_, unsigned int flags) for (networkName = networkNameList; networkName != NULL; networkName = networkName->_next) { if (STREQ(networkName->value, hostPortGroup->spec->name)) { - def->portGroups[def->nPortGroups].name = strdup(networkName->value); + portGroup = &def->portGroups[def->nPortGroups]; + ++def->nPortGroups; + + portGroup->name = strdup(networkName->value); - if (def->portGroups[def->nPortGroups].name == NULL) { + if (portGroup->name == NULL) { virReportOOMError(); goto cleanup; } @@ -834,13 +847,28 @@ esxNetworkGetXMLDesc(virNetworkPtr network_, unsigned int flags) if (hostPortGroup->spec->policy != NULL) { if (esxShapingPolicyToBandwidth (hostPortGroup->spec->policy->shapingPolicy, - &def->portGroups[def->nPortGroups].bandwidth) < 0) { + &portGroup->bandwidth) < 0) { ++def->nPortGroups; goto cleanup; } } - ++def->nPortGroups; + if (hostPortGroup->spec->vlanId->value > 0) { + if (hostPortGroup->spec->vlanId->value == 4095) { + portGroup->vlan.trunk = true; + } + + portGroup->vlan.nTags = 1; + + if (VIR_ALLOC_N(portGroup->vlan.tag, 1) < 0) { + virReportOOMError(); + goto cleanup; + } + + *portGroup->vlan.tag = + hostPortGroup->spec->vlanId->value; + } + break; } } -- 1.7.4.1

3 3

[libvirt] switching both libvirt and netcf packages to use libnl3 instead of libnl-1.1
by Laine Stump 24 Aug '12

24 Aug '12

Both libvirt and the netcf library it links to use libnl for doing things with netlink sockets. libnl comes in a couple of different and API-incompatible flavors, namely libnl-1.1 and libnl-3. For a long time Fedora had just libnl-1.1, and libvirt only supported libnl-1.1 Awhile back Fedora gained support for using libnl-3 instead of libnl-1.1, and did so in a way that allows both libraries (both runtime and -devel) to still exist on the system at the same time, and programs can pick which one they want to use. Even more recently libvirt and netcf got patches to support building with libnl-3 rather than libnl-1.1 (preferring libnl-3 if both are available). Since development (and even bugfixes) of libnl-1.1 is basically dead, I want to switch over the F18 and later builds of libvirt to use libnl-3, but this will take careful timing (and some restrictions in the libvirt specfile) because attempting to run a libnl3-linked libvirt with a netcf.so that was linked with libnl-1.1 (or vice versa) results in a segv. Other than the basic use of a Requires: line in the libvirt specfile to make the newly-libnl3-enabled build of libvirt require a similarly-enabled build of netcf, does anyone have any recommendations on making this transition as painless as possible (and as soon as possible - I obviously want to do this before F18 is released).

1 0

[libvirt] [PATCHv2 0/3] split snapshot conf out of domain conf
by Eric Blake 24 Aug '12

24 Aug '12

Rebased version from v1 [1], only changes are catering to other changes in domain_conf in the meantime, and (temporarily) dropping the XML changes of the original patch 4 until I have code that can actually use the XML in 0.10.1. I'm hoping these three patches still qualify for 0.10.0. [1] https://www.redhat.com/archives/libvir-list/2012-August/msg01250.html Eric Blake (3): snapshot: make virDomainSnapshotObjList opaque snapshot: split snapshot conf code into own file snapshot: rename an enum po/POTFILES.in | 1 + src/Makefile.am | 3 +- src/conf/domain_conf.c | 933 +------------------------------------------- src/conf/domain_conf.h | 143 +------ src/conf/snapshot_conf.c | 972 ++++++++++++++++++++++++++++++++++++++++++++++ src/conf/snapshot_conf.h | 157 ++++++++ src/esx/esx_driver.c | 1 + src/libvirt_private.syms | 5 +- src/qemu/qemu_command.c | 1 + src/qemu/qemu_domain.c | 7 +- src/qemu/qemu_domain.h | 3 +- src/qemu/qemu_driver.c | 69 ++-- src/qemu/qemu_migration.c | 2 +- src/vbox/vbox_tmpl.c | 1 + 14 files changed, 1209 insertions(+), 1089 deletions(-) create mode 100644 src/conf/snapshot_conf.c create mode 100644 src/conf/snapshot_conf.h -- 1.7.11.4

2 10

[libvirt] [PATCH] xen-xs: fix uuid of renamed domain
by Philipp Hahn 24 Aug '12

24 Aug '12

When the XenStore tdb lives persistently and is not cleared between host reboots, Xend (version 3.4 and 4.1) re-creates the domain information located in XenStore below /vm/$UUID. (According to the xen-3.2-commit hg265950e3df69 to fix a problem when locally migrating a domain to the host itself.) When doing so a version number is added to the UUID separated by one dash, which confuses xenStoreDomainIntroduced(): It iterates over all domains and tries to lookup all inactive domains using xenStoreDomainGetUUID(), which fails if the running domain is renamed: virUUIDParse() fails to parse the versioned UUID and the domain is flagged as missing. When this happens the function delays .2s and re-tries 20 times again, multiplied by the number of renamed VMs. 14:48:38.878: 4285: debug : xenStoreDomainIntroduced:1354 : Some domains were missing, trying again This adds a significant delay: # time virsh list >/dev/null real 0m6.529s # xenstore-list /vm 00000000-0000-0000-0000-000000000000 00000000-0000-0000-0000-000000000000-1 00000000-0000-0000-0000-000000000000-2 00000000-0000-0000-0000-000000000000-3 00000000-0000-0000-0000-000000000000-4 00000000-0000-0000-0000-000000000000-5 7c06121e-90c3-93d4-0126-50481d485cca 00000000-0000-0000-0000-000000000000-6 00000000-0000-0000-0000-000000000000-7 144ad19d-dfb4-2f80-8045-09196bb8784f 00000000-0000-0000-0000-000000000000-8 144ad19d-dfb4-2f80-8045-09196bb8784f-1 00000000-0000-0000-0000-000000000000-9 00000000-0000-0000-0000-000000000000-10 00000000-0000-0000-0000-000000000000-11 00000000-0000-0000-0000-000000000000-12 00000000-0000-0000-0000-000000000000-13 00000000-0000-0000-0000-000000000000-14 144ad19d-dfb4-2f80-8045-09196bb8784f-2 00000000-0000-0000-0000-000000000000-15 144ad19d-dfb4-2f80-8045-09196bb8784f-3 00000000-0000-0000-0000-000000000000-16 The patch adds truncation of the UUID as read from the XenStore path before passing it to virUUIDParse(). The same issue is reported at <http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=666135> Signed-off-by: Philipp Hahn <hahn(a)univention.de> --- src/xen/xs_internal.c | 5 ++++- 1 files changed, 4 insertions(+), 1 deletions(-) diff --git a/src/xen/xs_internal.c b/src/xen/xs_internal.c index e56d1a4..cdaef17 100644 --- a/src/xen/xs_internal.c +++ b/src/xen/xs_internal.c @@ -1115,8 +1115,11 @@ int xenStoreDomainGetUUID(virConnectPtr conn, snprintf(prop, 199, "/local/domain/%d/vm", id); prop[199] = 0; /* This will return something like - * /vm/00000000-0000-0000-0000-000000000000 */ + * /vm/00000000-0000-0000-0000-000000000000[-*] */ uuidstr = xs_read(priv->xshandle, 0, prop, &len); + /* Strip optional version suffix when VM was renamed */ + if (len > 40) /* strlen('/vm/') + VIR_UUID_STRING_BUFLEN - sizeof('\0') */ + uuidstr[40] = '\0'; /* remove "/vm/" */ ret = virUUIDParse(uuidstr + 4, uuid); -- 1.7.1

2 1

[libvirt] [PATCH 1/2] security: Add DAC to security_drivers
by Michal Privoznik 24 Aug '12

24 Aug '12

Currently, if users set 'security_driver="dac"' in qemu.conf libvirtd fails to initialize as DAC driver is not found because it is missing in our security drivers array. --- src/security/security_driver.c | 2 ++ 1 files changed, 2 insertions(+), 0 deletions(-) diff --git a/src/security/security_driver.c b/src/security/security_driver.c index f450a94..e6da220 100644 --- a/src/security/security_driver.c +++ b/src/security/security_driver.c @@ -35,6 +35,7 @@ # include "security_apparmor.h" #endif +#include "security_dac.h" #include "security_nop.h" #define VIR_FROM_THIS VIR_FROM_SECURITY @@ -46,6 +47,7 @@ static virSecurityDriverPtr security_drivers[] = { #ifdef WITH_SECDRIVER_APPARMOR &virAppArmorSecurityDriver, #endif + &virSecurityDriverDAC, &virSecurityDriverNop, /* Must always be last, since it will always probe */ }; -- 1.7.8.6

2 4

[libvirt] [PATCH] libssh2_transport: Add docs to remote.html
by Peter Krempa 24 Aug '12

24 Aug '12

Describe the existence of the transport driver and document the configurable options. --- docs/remote.html.in | 52 +++++++++++++++++++++++++++++++++++++++++++++++----- 1 file changed, 47 insertions(+), 5 deletions(-) diff --git a/docs/remote.html.in b/docs/remote.html.in index e6af4c2..40696b9 100644 --- a/docs/remote.html.in +++ b/docs/remote.html.in @@ -136,8 +136,14 @@ Remote libvirt supports a range of transports: <dd> Unencrypted TCP/IP socket. Not recommended for production use, this is normally disabled, but an administrator can enable it for testing or use over a trusted network. - The standard port is 16509. - </dd> + The standard port is 16509. </dd> + <dt> libssh2 </dt> + <dd> Transport over the SSH protocol using + <a href="http://libssh2.org/" title="libssh2 homepage">libssh2</a> instead +of the OpenSSH binary. This transport uses the libvirt authentication callback for +all ssh authentication calls and therefore supports keyboard-interactive authentication +even with graphical management applications. As with the classic ssh transport +netcat is required on the remote side.</dd> </dl> <p> The default transport, if no other is specified, is <code>tls</code>. @@ -182,6 +188,9 @@ Connect to a libvirtd daemon offering unencrypted TCP/IP connections on localhost port 5000 and use the test driver with default settings. </li> +<li><code>qemu+libssh2://user@host/system?known_hosts=/home/user/.ssh/known_hosts</code><br/> — +Connect to a remote host using a ssh connection with the libssh2 driver +and use a different known_hosts file.</li> </ul> <h4> <a name="Remote_URI_parameters">Extra parameters</a> @@ -237,7 +246,7 @@ Note that parameter values must be <td> <code>socket</code> </td> - <td> unix, ssh </td> + <td> unix, ssh, libssh2 </td> <td> The path to the Unix domain socket, which overrides the compiled-in default. For ssh transport, this is passed to @@ -252,7 +261,7 @@ Note that parameter values must be <td> <code>netcat</code> </td> - <td> ssh </td> + <td> ssh, libssh2 </td> <td> The name of the netcat command on the remote machine. The default is <code>nc</code>. For ssh transport, libvirt @@ -277,7 +286,7 @@ Note that parameter values must be <td> <code>keyfile</code> </td> - <td> ssh </td> + <td> ssh, libssh2 </td> <td> The name of the private key file to use to authentication to the remote machine. If this option is not used the default keys are used. @@ -341,6 +350,39 @@ Note that parameter values must be <td colspan="2"/> <td> Example: <code>pkipath=/tmp/pki/client</code> </td> </tr> + <tr> + <td> + <code>known_hosts</code> + </td> + <td> libssh2 </td> + <td> + Path to the known_hosts file to verify the host key agains. LibSSH2 + supports OpenSSH-style known_hosts files, although it does not support + all key types, so using files created by the OpenSSH binary may result + into truncating the known_hosts file. It's recommended to use the default + known_hosts file is located in libvirt's client local configuration + directory e.g.: ~/.config/libvirt/known_hosts. Note: Use absolute paths. +</td> + </tr> + <tr> + <td colspan="2"/> + <td> Example: <code>known_hosts=/root/.ssh/known_hosts</code> </td> + </tr> + <tr> + <td> + <code>sshauth</code> + </td> + <td> libssh2 </td> + <td> + A comma separated list of authentication methods to use. Default (is + "agent,privkey,keyboard-interactive". The order of the methods is perserved. + Some methods may require additional parameters. +</td> + </tr> + <tr> + <td colspan="2"/> + <td> Example: <code>sshauth=privkey,agent</code> </td> + </tr> </table> <h3> <a name="Remote_certificates">Generating TLS certificates</a> -- 1.7.12

2 2

[libvirt] How to solve this problem---libvirtError: this function is not supported by the connection driver: virNWFilterDefineXML
by Zhihong Wang 24 Aug '12

24 Aug '12

Dear all: I am trying to deploy openstack on xen hypervisor.Since XenSever or XCP is not available on our customized virtulization platform. We decide to use libvirt on xen as a middleware operating the xen hypervisor. But when I try to boot a instance , xen failed to response and I find the error in my nova.log as follows: 2012-08-24 09:17:25 ERROR nova.compute.manager [req-4b60e9a9-680c-457b-bf93-b71fb6230245 562096602c5e41748dd568d5db0f7553 6bd4a2580bcd4bef890d7ba4da82e925] [instance: 3645be21-9b5b-4718-9c9f-f6ba3588f7b4] Instance failed to spawn 2012-08-24 09:17:25 TRACE nova.compute.manager [instance: 3645be21-9b5b-4718-9c9f-f6ba3588f7b4] Traceback (most recent call last): 2012-08-24 09:17:25 TRACE nova.compute.manager [instance: 3645be21-9b5b-4718-9c9f-f6ba3588f7b4] File "/usr/lib/python2.6/site-packages/nova-2012.1-py2.6.egg/nova/compute/manager.py", line 592, in _spawn 2012-08-24 09:17:25 TRACE nova.compute.manager [instance: 3645be21-9b5b-4718-9c9f-f6ba3588f7b4] self._legacy_nw_info(network_info), block_device_info) 2012-08-24 09:17:25 TRACE nova.compute.manager [instance: 3645be21-9b5b-4718-9c9f-f6ba3588f7b4] File "/usr/lib/python2.6/site-packages/nova-2012.1-py2.6.egg/nova/exception.py", line 114, in wrapped 2012-08-24 09:17:25 TRACE nova.compute.manager [instance: 3645be21-9b5b-4718-9c9f-f6ba3588f7b4] return f(*args, **kw) 2012-08-24 09:17:25 TRACE nova.compute.manager [instance: 3645be21-9b5b-4718-9c9f-f6ba3588f7b4] File "/usr/lib/python2.6/site-packages/nova-2012.1-py2.6.egg/nova/virt/libvirt/connection.py", line 899, in spawn 2012-08-24 09:17:25 TRACE nova.compute.manager [instance: 3645be21-9b5b-4718-9c9f-f6ba3588f7b4] self.firewall_driver.setup_basic_filtering(instance, network_info) 2012-08-24 09:17:25 TRACE nova.compute.manager [instance: 3645be21-9b5b-4718-9c9f-f6ba3588f7b4] File "/usr/lib/python2.6/site-packages/nova-2012.1-py2.6.egg/nova/virt/libvirt/firewall.py", line 231, in setup_basic_filtering 2012-08-24 09:17:25 TRACE nova.compute.manager [instance: 3645be21-9b5b-4718-9c9f-f6ba3588f7b4] self.nwfilter.setup_basic_filtering(instance, network_info) 2012-08-24 09:17:25 TRACE nova.compute.manager [instance: 3645be21-9b5b-4718-9c9f-f6ba3588f7b4] File "/usr/lib/python2.6/site-packages/nova-2012.1-py2.6.egg/nova/virt/libvirt/firewall.py", line 102, in setup_basic_filtering 2012-08-24 09:17:25 TRACE nova.compute.manager [instance: 3645be21-9b5b-4718-9c9f-f6ba3588f7b4] self._ensure_static_filters() 2012-08-24 09:17:25 TRACE nova.compute.manager [instance: 3645be21-9b5b-4718-9c9f-f6ba3588f7b4] File "/usr/lib/python2.6/site-packages/nova-2012.1-py2.6.egg/nova/virt/libvirt/firewall.py", line 130, in _ensure_static_filters 2012-08-24 09:17:25 TRACE nova.compute.manager [instance: 3645be21-9b5b-4718-9c9f-f6ba3588f7b4] 'allow-dhcp-server'])) 2012-08-24 09:17:25 TRACE nova.compute.manager [instance: 3645be21-9b5b-4718-9c9f-f6ba3588f7b4] File "/usr/lib/python2.6/site-packages/nova-2012.1-py2.6.egg/nova/virt/libvirt/firewall.py", line 150, in _define_filter 2012-08-24 09:17:25 TRACE nova.compute.manager [instance: 3645be21-9b5b-4718-9c9f-f6ba3588f7b4] tpool.execute(self._conn.nwfilterDefineXML, xml) 2012-08-24 09:17:25 TRACE nova.compute.manager [instance: 3645be21-9b5b-4718-9c9f-f6ba3588f7b4] File "/usr/lib/python2.6/site-packages/eventlet/tpool.py", line 76, in tworker 2012-08-24 09:17:25 TRACE nova.compute.manager [instance: 3645be21-9b5b-4718-9c9f-f6ba3588f7b4] rv = meth(*args,**kwargs) 2012-08-24 09:17:25 TRACE nova.compute.manager [instance: 3645be21-9b5b-4718-9c9f-f6ba3588f7b4] File "/usr/lib64/python2.6/site-packages/libvirt.py", line 2628, in nwfilterDefineXML 2012-08-24 09:17:25 TRACE nova.compute.manager [instance: 3645be21-9b5b-4718-9c9f-f6ba3588f7b4] if ret is None:raise libvirtError('virNWFilterDefineXML() failed', conn=self) 2012-08-24 09:17:25 TRACE nova.compute.manager [instance: 3645be21-9b5b-4718-9c9f-f6ba3588f7b4] libvirtError: this function is not supported by the connection driver: virNWFilterDefineXML 2012-08-24 09:17:25 TRACE nova.compute.manager [instance: 3645be21-9b5b-4718-9c9f-f6ba3588f7b4] It seems that libvirt doesn't support virNWFilterDefineXML on xen,Is there any good solutions? Or how much work should be done to add this function to libvirt? I would take it into consideration seriously if I could fix this problem

2 1

[libvirt] [PATCH] openvz: check the exitstatus of vzlist
by Laine Stump 24 Aug '12

24 Aug '12

I noticed this while auditing all calls to virCommandRun that request an exit status from virCommandRun. Two functions in the openvz driver openvzDomainGetBarrierLimit openvzDomainSetBarrierLimit request an exit status from virCommandRun (thus assuring that virCommandRun won't log any errors just due to a non-0 exit status), but then fail to examine that exit status. This could result in the functions believing that the call to "vzlist" was successful, even though it may have encountered an error. --- src/openvz/openvz_driver.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/src/openvz/openvz_driver.c b/src/openvz/openvz_driver.c index 8257ed5..a1d3b42 100644 --- a/src/openvz/openvz_driver.c +++ b/src/openvz/openvz_driver.c @@ -1707,7 +1707,7 @@ openvzDomainGetBarrierLimit(virDomainPtr domain, virCommandSetOutputBuffer(cmd, &output); virCommandAddArgFormat(cmd, "-o%s.b,%s.l", param, param); virCommandAddArg(cmd, domain->name); - if (virCommandRun(cmd, &status)) { + if (virCommandRun(cmd, &status) < 0 || status != 0) { virReportError(VIR_ERR_OPERATION_FAILED, _("Failed to get %s for %s: %d"), param, domain->name, status); @@ -1758,7 +1758,7 @@ openvzDomainSetBarrierLimit(virDomainPtr domain, virCommandAddArgFormat(cmd, "--%s", param); virCommandAddArgFormat(cmd, "%llu:%llu", barrier, limit); virCommandAddArg(cmd, "--save"); - if (virCommandRun(cmd, &status)) { + if (virCommandRun(cmd, &status) < 0 || status != 0) { virReportError(VIR_ERR_OPERATION_FAILED, _("Failed to set %s for %s: %d"), param, domain->name, status); -- 1.7.11.4

2 2

[libvirt] [PATCH] nwfilter: don't log error if firewalld is disabled
by Laine Stump 24 Aug '12

24 Aug '12

The original patch to support firewalld in nwfilter wasn't personally checking the exit status of firewall-cmd, but was instead sending NULL in the *exitstatus arg, which meant that virCommandWait would log an error just for the exit status being non-0 (and a "more scary than useful" error at that). We don't want to treat this as an error, though, just as a reason to use standard (ip|eb)tables commands instead of firewall-cmd. This patch modifies the virCommandRun in the nwfilter code to request status back from the caller. This avoids virCommandWait logging an error message, and allows the caller to do as it likes after examining the status. --- src/nwfilter/nwfilter_ebiptables_driver.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/src/nwfilter/nwfilter_ebiptables_driver.c b/src/nwfilter/nwfilter_ebiptables_driver.c index b008879..8f7a453 100644 --- a/src/nwfilter/nwfilter_ebiptables_driver.c +++ b/src/nwfilter/nwfilter_ebiptables_driver.c @@ -4140,6 +4140,7 @@ ebiptablesDriverInitWithFirewallD(void) virBuffer buf = VIR_BUFFER_INITIALIZER; char *firewall_cmd_path; char *output = NULL; + int status; int ret = -1; if (!virNWFilterDriverIsWatchingFirewallD()) @@ -4155,8 +4156,8 @@ ebiptablesDriverInitWithFirewallD(void) "%s", CMD_STOPONERR(1)); - if (ebiptablesExecCLI(&buf, NULL, &output) == 0 && - strlen(output) == 0) { + if (ebiptablesExecCLI(&buf, &status, &output) == 0 && + status == 0) { VIR_DEBUG("Using firewall-cmd in nwfilter_ebiptables_driver."); ignore_value(virAsprintf(&ebtables_cmd_path, -- 1.7.11.4

2 3

[libvirt] Guest log getting written in cache directory rather /var/log/libvirt/qemu
by Richa Marwaha 23 Aug '12

23 Aug '12

Hi Everyone I am running libvirt with selinux driver under unprivileged user and the Guest.log file is not written into /var/log/libvirt/qemu/Guest.log but remains in Home Directory/.cache/libvirt/qemu/log/Guest.log. Does anyone know why its not writing to the required log location. Thanks Regards, Richa Marwaha LTC Security Email address:rmarwah@us.ibm.com

2 4

[libvirt] [PATCH] network: fix virtual network bridge delay setting
by Laine Stump 23 Aug '12

23 Aug '12

libvirt's network config documents that a bridge's STP "forward delay" (called "delay" in the XML) should be specified in seconds, but virNetDevBridgeSetSTPDelay() assumes that it is given a delay in milliseconds (although the comment at the top of the function incorrectly says "seconds". This fixes the comment, and converts the delay to milliseconds before calling virNetDevBridgeSetSTPDelay(). --- (Fortunately the default delay (and what most people set) is 0) src/network/bridge_driver.c | 6 +++++- src/util/virnetdevbridge.c | 4 ++-- 2 files changed, 7 insertions(+), 3 deletions(-) diff --git a/src/network/bridge_driver.c b/src/network/bridge_driver.c index 59b2c10..53eebed 100644 --- a/src/network/bridge_driver.c +++ b/src/network/bridge_driver.c @@ -1826,8 +1826,12 @@ networkStartNetworkVirtual(struct network_driver *driver, } /* Set bridge options */ + + /* delay is configured in seconds, but virNetDevBridgeSetSTPDelay + * expects milliseconds + */ if (virNetDevBridgeSetSTPDelay(network->def->bridge, - network->def->delay) < 0) + network->def->delay * 1000) < 0) goto err1; if (virNetDevBridgeSetSTP(network->def->bridge, diff --git a/src/util/virnetdevbridge.c b/src/util/virnetdevbridge.c index a616d8e..7b11bee 100644 --- a/src/util/virnetdevbridge.c +++ b/src/util/virnetdevbridge.c @@ -1,5 +1,5 @@ /* - * Copyright (C) 2007-2011 Red Hat, Inc. + * Copyright (C) 2007-2012 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 @@ -378,7 +378,7 @@ int virNetDevBridgeRemovePort(const char *brname, /** * virNetDevBridgeSetSTPDelay: * @brname: the bridge name - * @delay: delay in seconds + * @delay: delay in milliseconds * * Set the bridge forward delay * -- 1.7.11.4

2 2

[libvirt] [PATCH v4 0/3] Per-guest S3/S4 configuration
by Martin Kletzander 23 Aug '12

23 Aug '12

There is a way to tell qemu how (not) to advertise S3/S4 ACPI sleep state capability to the guest OS. This series includes the capability to set this in the XML and also covers all the handling from qemu point of view including checking for the support, parameter parsing, command building, checking before sending the signals through guest agent and also tests and documentation. -- v4: - Changed attributes to elements - Merged documentation with XML extension - Fixed error reporting and other tweaks Michal pointed out v3: - Option names are change according to Eric and Doug - Added docs (formatdomain) v2: - Modified the patch to reflect danpb's notes (according to qemu people the setting the disable_s[34] parameter to 0/1 ensures that the states will be enabled/disabled respectively) *** BLURB HERE *** Martin Kletzander (3): Add per-guest S3/S4 state configuration qemu: Add support for S3/S4 state configuration tests: Add tests for qemu S3/S4 state configuration docs/formatdomain.html.in | 24 +++++++++ docs/schemas/domaincommon.rng | 39 ++++++++++++++ src/conf/domain_conf.c | 52 +++++++++++++++++- src/conf/domain_conf.h | 15 ++++++ src/libvirt_private.syms | 2 + src/qemu/qemu_capabilities.c | 7 +++ src/qemu/qemu_capabilities.h | 2 + src/qemu/qemu_command.c | 62 ++++++++++++++++++++++ src/qemu/qemu_driver.c | 17 ++++++ tests/qemuargv2xmltest.c | 3 ++ .../qemuxml2argv-misc-disable-s3.args | 4 ++ .../qemuxml2argv-misc-disable-s3.xml | 29 ++++++++++ .../qemuxml2argv-misc-disable-suspends.args | 4 ++ .../qemuxml2argv-misc-disable-suspends.xml | 30 +++++++++++ .../qemuxml2argv-misc-enable-s4.args | 4 ++ .../qemuxml2argv-misc-enable-s4.xml | 29 ++++++++++ tests/qemuxml2argvtest.c | 4 ++ tests/qemuxml2xmltest.c | 3 ++ 18 files changed, 329 insertions(+), 1 deletion(-) create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-misc-disable-s3.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-misc-disable-s3.xml create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-misc-disable-suspends.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-misc-disable-suspends.xml create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-misc-enable-s4.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-misc-enable-s4.xml -- 1.7.12

3 9

[libvirt] [PATCH] qemu: forbid suspend if already pmsuspended
by Michal Privoznik 23 Aug '12

23 Aug '12

If a domain is pmsuspended then virsh suspend will succeed. Beside obvious flaw, virsh resume will report success and change domain state to running which is another mistake. Therefore we must forbid any attempts for suspend and resume when pmsuspended. --- src/libvirt.c | 4 ++++ src/qemu/qemu_driver.c | 18 ++++++++++++++++-- 2 files changed, 20 insertions(+), 2 deletions(-) diff --git a/src/libvirt.c b/src/libvirt.c index e0ac391..b034ed6 100644 --- a/src/libvirt.c +++ b/src/libvirt.c @@ -2329,6 +2329,8 @@ virDomainRef(virDomainPtr domain) * hypervisor level will stay allocated. Use virDomainResume() to reactivate * the domain. * This function may require privileged access. + * Moreover, suspend may not be supported if domain is in some + * special state like VIR_DOMAIN_PMSUSPENDED. * * Returns 0 in case of success and -1 in case of failure. */ @@ -2375,6 +2377,8 @@ error: * Resume a suspended domain, the process is restarted from the state where * it was frozen by calling virDomainSuspend(). * This function may require privileged access + * Moreover, resume may not be supported if domain is in some + * special state like VIR_DOMAIN_PMSUSPENDED. * * Returns 0 in case of success and -1 in case of failure. */ diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 8c7b34d..256fa71 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -1572,6 +1572,7 @@ static int qemudDomainSuspend(virDomainPtr dom) { qemuDomainObjPrivatePtr priv; virDomainPausedReason reason; int eventDetail; + int state; qemuDriverLock(driver); vm = virDomainFindByUUID(&driver->domains, dom->uuid); @@ -1607,7 +1608,13 @@ static int qemudDomainSuspend(virDomainPtr dom) { "%s", _("domain is not running")); goto endjob; } - if (virDomainObjGetState(vm, NULL) != VIR_DOMAIN_PAUSED) { + + state = virDomainObjGetState(vm, NULL); + if (state == VIR_DOMAIN_PMSUSPENDED) { + virReportError(VIR_ERR_OPERATION_INVALID, + "%s", _("domain is pmsuspended")); + goto endjob; + } else if (state != VIR_DOMAIN_PAUSED) { if (qemuProcessStopCPUs(driver, vm, reason, QEMU_ASYNC_JOB_NONE) < 0) { goto endjob; } @@ -1639,6 +1646,7 @@ static int qemudDomainResume(virDomainPtr dom) { virDomainObjPtr vm; int ret = -1; virDomainEventPtr event = NULL; + int state; qemuDriverLock(driver); vm = virDomainFindByUUID(&driver->domains, dom->uuid); @@ -1659,7 +1667,13 @@ static int qemudDomainResume(virDomainPtr dom) { "%s", _("domain is not running")); goto endjob; } - if (virDomainObjGetState(vm, NULL) == VIR_DOMAIN_PAUSED) { + + state = virDomainObjGetState(vm, NULL); + if (state == VIR_DOMAIN_PMSUSPENDED) { + virReportError(VIR_ERR_OPERATION_INVALID, + "%s", _("domain is pmsuspended")); + goto endjob; + } else if (state == VIR_DOMAIN_PAUSED) { if (qemuProcessStartCPUs(driver, vm, dom->conn, VIR_DOMAIN_RUNNING_UNPAUSED, QEMU_ASYNC_JOB_NONE) < 0) { -- 1.7.8.6

2 2