January 2025 - Devel - libvirt List Archives

[PATCH] access: fix po_check when polkit is disabled
by Daniel P. Berrangé 13 Jan '25

13 Jan '25

The generated org.libvirt.api.policy.in file was recently added to the POTFILES list as it contains translatable messages. It is only generated when WITH_POLKIT && WITH_LIBVIRTD is satisfied though, resulting in the 'po_check' syntax rule failing if either of those conditions are not met. It is harmless to unconditionally generate this file, as a separate rule takes care of of installing it, and the latter remains under the build conditions. Signed-off-by: Daniel P. Berrangé <berrange(a)redhat.com> --- src/access/meson.build | 21 ++++++++++++--------- 1 file changed, 12 insertions(+), 9 deletions(-) diff --git a/src/access/meson.build b/src/access/meson.build index 0cc9cf2b79..d1336c0093 100644 --- a/src/access/meson.build +++ b/src/access/meson.build @@ -63,19 +63,22 @@ foreach name : [ 'remote', 'qemu', 'lxc' ] ) endforeach +# Generated outside the WITH_POLKIT / WITH_LIBVIRTD condition +# because syntax-check/po_check requires this to exist +polgen = custom_target( + 'org.libvirt.api.policy.in', + input: access_perm_h, + output: 'org.libvirt.api.policy.in', + command: [ meson_python_prog, python3_prog, genpolkit_prog, '@INPUT@' ], + capture: true, + build_by_default: true, +) +access_generated += polgen + if conf.has('WITH_POLKIT') access_sources += access_polkit_sources if conf.has('WITH_LIBVIRTD') - polgen = custom_target( - 'org.libvirt.api.policy.in', - input: access_perm_h, - output: 'org.libvirt.api.policy.in', - command: [ meson_python_prog, python3_prog, genpolkit_prog, '@INPUT@' ], - capture: true, - build_by_default: true, - ) - access_generated += polgen i18n.merge_file( input: polgen, -- 2.47.1

2 1

[PATCH] qemu: support vlan change for linux host bridge during update-device
by Laine Stump 13 Jan '25

13 Jan '25

Since we previously only supported vlan tagging for interfaces connected to an OVS bridge [*], the code in qemuChangeNet() (used by the update-device API) assumed an interface with modified vlan config was on an OVS bridge, and would call the OVS-specific virNetDevOpenvswitchUpdateVlan(). Now that we support vlan tagging for interfaces connected to a standard Linux host bridge, we must check the type of connection and only call the OVS function when connected to an OVS bridge *both before and after the update*. Otherwise we just set the flag to re-connect to the bridge, which has the side effect of redoing the vlan setup. ([*] or an SRIOV VF assigned using VFIO, but we don't support *any runtime changes to that type of netdev so it's irrelevant here.) Signed-off-by: Laine Stump <laine(a)redhat.com> --- I'm fine with this going in either before or after the 11.0.0 release (would prefer before since it is a bugfix), but if it's going into the release I'd rather it be included with RC2 rather than only in the final release. For that reason, if someone ACKs the patch and agrees that it's safe to go in, please push it for me (since Jiri will almost certainly tag RC2 before I'm conscious on Monday morning) src/qemu/qemu_hotplug.c | 32 ++++++++++++++++++++++++-------- 1 file changed, 24 insertions(+), 8 deletions(-) diff --git a/src/qemu/qemu_hotplug.c b/src/qemu/qemu_hotplug.c index de0777d330..5a7e6c3b12 100644 --- a/src/qemu/qemu_hotplug.c +++ b/src/qemu/qemu_hotplug.c @@ -4144,8 +4144,13 @@ qemuDomainChangeNet(virQEMUDriver *driver, * they don't apply to a particular type. */ - if (!virNetDevVlanEqual(virDomainNetGetActualVlan(olddev), - virDomainNetGetActualVlan(newdev))) { + /* since attaching to a new bridge will re-do the vlan setup, + * we don't need to separately do that in the case that we're + * already switching to a different bridge + */ + if (!(needBridgeChange || + virNetDevVlanEqual(virDomainNetGetActualVlan(olddev), + virDomainNetGetActualVlan(newdev)))) { needVlanUpdate = true; } @@ -4215,6 +4220,23 @@ qemuDomainChangeNet(virQEMUDriver *driver, needReplaceDevDef = true; } + if (needVlanUpdate) { + if (virDomainNetDefIsOvsport(olddev) && virDomainNetDefIsOvsport(newdev)) { + /* optimization if we're attached to an OVS bridge. This + * will redo vlan setup without needing to re-attach the + * tap device to the bridge + */ + if (virNetDevOpenvswitchUpdateVlan(newdev->ifname, &newdev->vlan) < 0) + goto cleanup; + } else { + /* vlan setup is done as a part of reconnecting the tap + * device to a new bridge (either OVS or Linux host bridge). + */ + needBridgeChange = true; + } + needReplaceDevDef = true; + } + if (needBridgeChange) { if (qemuDomainChangeNetBridge(vm, olddev, newdev) < 0) goto cleanup; @@ -4266,12 +4288,6 @@ qemuDomainChangeNet(virQEMUDriver *driver, goto cleanup; } - if (needVlanUpdate) { - if (virNetDevOpenvswitchUpdateVlan(newdev->ifname, &newdev->vlan) < 0) - goto cleanup; - needReplaceDevDef = true; - } - if (needReplaceDevDef) { /* the changes above warrant replacing olddev with newdev in * the domain's nets list. -- 2.47.1

2 1

[PATCH 0/2] news: add a couple items before release
by Laine Stump 13 Jan '25

13 Jan '25

This adds a "new feature" for the VLAN tagging support on Linux host bridges added by Leigh Brown, and an "improvement" (migration between mdev<->SRIOV vGPUs) by me. Feel free to push this for me so it can be included in RC2. *** BLURB HERE *** Laine Stump (2): news: note addition of vlan tagging support for Linux host bridges news: document support for mdev <-> SRIOV VF vGPU live migration NEWS.rst | 13 +++++++++++++ 1 file changed, 13 insertions(+) -- 2.47.1

2 3

[PATCH] qemu_migration: Do not consider post-copy active in postcopy-recover
by Jiri Denemark 13 Jan '25

13 Jan '25

The postcopy-recover migration state in QEMU means a connection for the migration stream was established. Depending on the schedulers on both hosts a relative timing of the corresponding MIGRATION event on the source host and the destination host may differ. Specifically it's possible that the source sees postcopy-recover while the destination is still in postcopy-paused. Currently the Perform phase on the source host ends when we get postcopy-recover event and the Finish phase on the destination host is called. If this is fast enough we can still see postcopy-paused state when the Finish phase starts waiting for migration to complete. This is interpreted as a failure and reported back to the caller. Even though the recovery may actually start just a few moments later. To avoid this race we now don't consider post-copy migration active in postcopy-recover state and keep waiting for postcopy-active event (in the success path). Thus the Finish phase is entered only after the migration switches to postcopy-active. In this state QEMU guarantees the destination already switched at least to postcopy-recover and we won't be confused be seeing an old postcopy-failed state. https://issues.redhat.com/browse/RHEL-73085 Signed-off-by: Jiri Denemark <jdenemar(a)redhat.com> --- src/qemu/qemu_migration.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/qemu/qemu_migration.c b/src/qemu/qemu_migration.c index 50e350b0c4..1582a738a3 100644 --- a/src/qemu/qemu_migration.c +++ b/src/qemu/qemu_migration.c @@ -1872,11 +1872,11 @@ qemuMigrationUpdateJobType(virDomainJobData *jobData) switch ((qemuMonitorMigrationStatus) priv->stats.mig.status) { case QEMU_MONITOR_MIGRATION_STATUS_POSTCOPY: - case QEMU_MONITOR_MIGRATION_STATUS_POSTCOPY_RECOVER: jobData->status = VIR_DOMAIN_JOB_STATUS_POSTCOPY; break; case QEMU_MONITOR_MIGRATION_STATUS_POSTCOPY_RECOVER_SETUP: + case QEMU_MONITOR_MIGRATION_STATUS_POSTCOPY_RECOVER: jobData->status = VIR_DOMAIN_JOB_STATUS_POSTCOPY_RECOVER; break; -- 2.47.1

2 1

[PATCH] conf: Do not parse hyperv features with passthrough mode
by Martin Kletzander 13 Jan '25

13 Jan '25

The schema does not allow that anyway and we then format them all back which leads to libvirt producing an invalid XML. Resolves: https://issues.redhat.com/browse/RHEL-70656 Signed-off-by: Martin Kletzander <mkletzan(a)redhat.com> --- src/conf/domain_conf.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index e658b68c48b6..3f88a77a8fea 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -16643,6 +16643,9 @@ virDomainFeaturesHyperVDefParse(virDomainDef *def, def->features[VIR_DOMAIN_FEATURE_HYPERV] = mode; + if (mode == VIR_DOMAIN_HYPERV_MODE_PASSTHROUGH) + return 0; + node = xmlFirstElementChild(node); while (node != NULL) { int feature; -- 2.47.1

2 1

Virtual manager error
by Domenico Panella 13 Jan '25

13 Jan '25

Hi everyone, when i try to run a VM through virt-manager application (Slackware 15), I get the following error: *Errore nell'avvio del dominio: internal error: unknown architecture 'loongarch64'** ** **Traceback (most recent call last):** ** File "/usr/share/virt-manager/virtManager/asyncjob.py", line 72, in cb_wrapper** ** callback(asyncjob, *args, **kwargs)** ** File "/usr/share/virt-manager/virtManager/asyncjob.py", line 108, in tmpcb** ** callback(*args, **kwargs)** ** File "/usr/share/virt-manager/virtManager/object/libvirtobject.py", line 57, in newfn** ** ret = fn(self, *args, **kwargs)** ** File "/usr/share/virt-manager/virtManager/object/domain.py", line 1402, in startup** ** self._backend.create()** ** File "/usr/lib64/python3.9/site-packages/libvirt.py", line 1373, in create** ** raise libvirtError('virDomainCreate() failed')** **libvirt.libvirtError: internal error: unknown architecture 'loongarch64'* How can i fix it? Cordially Domenico

2 1

[PATCH v4 0/4] iproute2 bridge vlan support
by Leigh Brown 11 Jan '25

11 Jan '25

I have incorporated Laine's feedback and added Reviewed-by tags. I also noticed that I hadn't updated the patch series description, so I updated it to reflect the change to using netlink, for posterity. Description ----------- The iproute2 bridge command supports the capability for VLAN filtering that allows each interface connected to a standard linux bridge to be configured to use one or more VLANs. For simple setups, this capability is enough to allow virtual machines or containers to be put onto separate VLANs without creating multiple bridges and VLANs on the host. The first patch adds a new function virNetDevBridgeVlanFilterSet that allows a VLAN filter to be added or removed from an interface associated with a bridge. The second patch adds virNetDevBridgeSetupVlans that will, given a virNetDevVlan structure, call virNetDevBridgeVlanFilterSet to apply the required VLAN filtering for the given interface. The third patch adjusts the domain and network validation to permit standard linux bridges to allow VLAN configuration and updates calls to virNetDevBridgeAddPort to pass the VLAN configuration. The fourth patch updates documentation to match the new capability. Changes since v3 ---------------- - Update patch series description. - Fix coding style. - Add G_GNUC_UNUSED annotation to virNetDevBridgeAddPort for MacOS. Changes since v2 ---------------- - Convert to use netlink rather than executing bridge vlan commands. - Add unsupported on this platform error message on FreeBSD. Changes since v1 ---------------- - Fix bug in virNetDevSetupVlans where bridge port has no native vlan. - Update bridge network validation to permit vlan configuration. - Update documentation to match the functionality. - Tweak some of the commit descriptions for clarity. Usage example ------------- Configure the host with systemd-networkd as follows: /etc/systemd/network/br0.netdev (br0.network not shown) [NetDev] Name=br0 Kind=bridge MACAddress=xx:xx:xx:xx:xx:xx [Bridge] VLANFiltering=on /etc/systemd/network/eno1.network [Match] Name=eno1 [Network] Bridge=br0 [Link] MTUBytes=9000 [BridgeVLAN] VLAN=40 [BridgeVLAN] VLAN=60 Then add <vlan> tags into the lxc or qemu config: lxc interface definition: <interface type='bridge'> <mac address='xx:xx:xx:xx:xx:xx'/> <source bridge='br0'/> <vlan> <tag id='40'/> </vlan> </interface> qemu interface definition: <interface type='network'> <mac address='xx:xx:xx:xx:xx:xx'/> <source network='br0'/> <vlan> <tag id='60'/> </vlan> <model type='virtio'/> <address type='pci' domain='0x0000' bus='0x01' slot='0x00' function='0x0'/> </interface> Then, after starting them, you will see the following $ sudo bridge vlan port vlan-id eno1 1 PVID Egress Untagged 40 60 br0 1 PVID Egress Untagged vnet0 60 PVID Egress Untagged vnet1 40 PVID Egress Untagged Regards, Leigh Brown (4): util: add netlink bridge vlan filtering util: Add vlan support to virNetDevBridgeAddPort Enable vlan support for standard linux bridges docs: standard linux bridges now support vlans docs/formatdomain.rst | 37 +++++++++--------- docs/formatnetwork.rst | 45 +++++++++++----------- src/conf/domain_validate.c | 3 +- src/lxc/lxc_process.c | 2 +- src/network/bridge_driver.c | 13 ++++--- src/util/virnetdevbridge.c | 75 +++++++++++++++++++++++++++++++++++-- src/util/virnetdevbridge.h | 4 +- src/util/virnetdevtap.c | 2 +- src/util/virnetlink.c | 66 ++++++++++++++++++++++++++++++++ src/util/virnetlink.h | 7 ++++ 10 files changed, 201 insertions(+), 53 deletions(-) -- 2.39.5

3 8

[libvirt PATCH] qemu: snapshot: delete disk image only if parent snapshot is external
by Pavel Hrdina 10 Jan '25

10 Jan '25

When we are deleting external snapshot that is not active we only need to delete overlay disk image of the parent snapshot. This works correctly even if parent snapshot is external and active as it will have another overlay created when user reverted to that snapshot. In case the parent snapshot is internal there are no overlay disk images created as everything is stored internally within the disk image. In this case we would delete the actual disk image storing internal snapshots and most likely the original disk image as well resulting in data loss once the VM is shutoff. Fixes: https://gitlab.com/libvirt/libvirt/-/issues/734 Signed-off-by: Pavel Hrdina <phrdina(a)redhat.com> --- src/qemu/qemu_snapshot.c | 14 ++++++++------ 1 file changed, 8 insertions(+), 6 deletions(-) diff --git a/src/qemu/qemu_snapshot.c b/src/qemu/qemu_snapshot.c index 18b2e478f6..80cd54bf33 100644 --- a/src/qemu/qemu_snapshot.c +++ b/src/qemu/qemu_snapshot.c @@ -3144,6 +3144,8 @@ qemuSnapshotDeleteExternalPrepareData(virDomainObj *vm, return -1; } + data->parentSnap = qemuSnapshotFindParentSnapForDisk(snap, data->snapDisk); + if (data->merge) { virStorageSource *snapDiskSrc = NULL; @@ -3185,8 +3187,6 @@ qemuSnapshotDeleteExternalPrepareData(virDomainObj *vm, qemuSnapshotGetDisksWithBackingStore(vm, snap, data); } - data->parentSnap = qemuSnapshotFindParentSnapForDisk(snap, data->snapDisk); - if (data->parentSnap && !virDomainSnapshotIsExternal(data->parentSnap)) { virReportError(VIR_ERR_OPERATION_UNSUPPORTED, "%s", _("deleting external snapshot that has internal snapshot as parent not supported")); @@ -3642,10 +3642,12 @@ qemuSnapshotDiscardExternal(virDomainObj *vm, if (!data->job) goto error; } else { - if (virStorageSourceInit(data->parentDomDisk->src) < 0 || - virStorageSourceUnlink(data->parentDomDisk->src) < 0) { - VIR_WARN("Failed to remove snapshot image '%s'", - data->snapDisk->name); + if (data->parentSnap && virDomainSnapshotIsExternal(data->parentSnap)) { + if (virStorageSourceInit(data->parentDomDisk->src) < 0 || + virStorageSourceUnlink(data->parentDomDisk->src) < 0) { + VIR_WARN("Failed to remove snapshot image '%s'", + data->snapDisk->name); + } } } } -- 2.47.1

2 2

[PATCH 0/1] RFC: Add Arm CCA support for getting capability information and running Realm VM
by Akio Kakuno 10 Jan '25

10 Jan '25

Hi, all. - This patch adds Arm CCA support to qemu driver for aarch64 system. CCA is an abbreviation for Arm Confidential Compute Architecture feature, it enhances the virtualization capabilities of the platform by separating the management of resources from access to those resources. - We are not yet at the stage where we can merge this patch as host linux/qemu suppor is no yet merged, but I would like to receive reviews and comments on the overall direction. [summary] - At this stage, all you can do is getting the CCA capability with the virsh domcapabilities command and start the CCA VM with the virsh create command. - capability info uses qemu QMP to query qemu options. The option that exists now is for selecting a hash algorithm. [Capability example] - Execution results of 'virsh domcapability" on qemu <domaincapabilities> ... <features> ... </sgx> <cca supported='yes'> <enum name='measurement-algo'> <value>sha256</value> <value>sha512</value> </enum> </cca> <hyperv supported='yes'> ... </features> </domaincapabilities> [XML example] <domain> ... <launchsecurity type='cca'> <measurement-algo>sha256</measurement-algo> </launchsecurity> ... </domain> [limitations/tests] - To obtain capability info, it is necessary to support the qemu QMP command, which qemu does not yet support. Therefore, I put a hack in the code at hand and only confirmed the communication. Also, I think we should check whether CPUFW supports CCA or not in qemu_firmware.c, but it is not yet implemented. - Verified that the CCA VM can be started from virsh create command. [software version] - I followed the steps in Linaro's blog below. https://linaro.atlassian.net/wiki/spaces/QEMU/pages/29051027459/Building+an… - The Qemu used was based on Linaro's qemu(9.1.91). https://git.codelinaro.org/linaro/dcap/qemu/-/tree/cca/v3?ref_type=heads Signed-off-by: Akio Kakuno <fj3333bs(a)fujitsu.com> Best Regards. Akio Kakuno (1): RFC: Add Arm CCA support for getting capability information and running Realm VM docs/formatdomain.rst | 28 ++++++ docs/formatdomaincaps.rst | 26 ++++- src/conf/domain_capabilities.c | 41 ++++++++ src/conf/domain_capabilities.h | 12 +++ src/conf/domain_conf.c | 13 +++ src/conf/domain_conf.h | 7 ++ src/conf/schemas/domaincaps.rng | 14 +++ src/conf/schemas/domaincommon.rng | 14 +++ src/conf/virconftypes.h | 2 + src/libvirt_private.syms | 1 + src/qemu/qemu_capabilities.c | 156 ++++++++++++++++++++++++++++++ src/qemu/qemu_capabilities.h | 4 + src/qemu/qemu_cgroup.c | 2 + src/qemu/qemu_command.c | 32 ++++++ src/qemu/qemu_driver.c | 2 + src/qemu/qemu_monitor.c | 10 ++ src/qemu/qemu_monitor.h | 3 + src/qemu/qemu_monitor_json.c | 104 ++++++++++++++++++++ src/qemu/qemu_monitor_json.h | 4 + src/qemu/qemu_namespace.c | 2 + src/qemu/qemu_process.c | 4 + src/qemu/qemu_validate.c | 7 ++ 22 files changed, 487 insertions(+), 1 deletion(-) -- 2.34.1

3 3

[PATCH] Passing a domain XML to a polkit rule engine
by Takuya Nakaike 09 Jan '25

09 Jan '25

This patch is a draft implementation to pass a domain XML to a polkit access driver. With this new feature, a polkit rule can verify the domain XML to be deployed on a host, and thus protect deploying a malicious VM. There is a discussion about this new feature in the following issue. https://gitlab.com/libvirt/libvirt/-/issues/719 Any question, comment, and suggestion are welcome. Thanks, Signed-off-by: Takuya Nakaike <nakaike(a)jp.ibm.com> --- src/access/viraccessdriverpolkit.c | 2 ++ src/access/viraccessmanager.c | 22 ++++++++++++++++++++++ src/access/viraccessmanager.h | 3 +++ src/libvirt_private.syms | 1 + src/qemu/qemu_driver.c | 3 +++ 5 files changed, 31 insertions(+) diff --git a/src/access/viraccessdriverpolkit.c b/src/access/viraccessdriverpolkit.c index 83381183a5..56457010e0 100644 --- a/src/access/viraccessdriverpolkit.c +++ b/src/access/viraccessdriverpolkit.c @@ -177,10 +177,12 @@ virAccessDriverPolkitCheckDomain(virAccessManager *manager, virAccessPermDomain perm) { char uuidstr[VIR_UUID_STRING_BUFLEN]; + char *xml = virAccessManagerGetXMLDesc(domain); const char *attrs[] = { "connect_driver", driverName, "domain_name", domain->name, "domain_uuid", uuidstr, + "xml", xml != NULL ? xml : "", NULL, }; virUUIDFormat(domain->uuid, uuidstr); diff --git a/src/access/viraccessmanager.c b/src/access/viraccessmanager.c index 6d9fdee5f1..7be31eb19a 100644 --- a/src/access/viraccessmanager.c +++ b/src/access/viraccessmanager.c @@ -44,11 +44,13 @@ struct _virAccessManager { virObjectLockable parent; virAccessDriver *drv; + virDomainXMLOption *xmlopt; void *privateData; }; static virClass *virAccessManagerClass; static virAccessManager *virAccessManagerDefault; +static virDomainXMLOption *virAccessManagerXMLOption = NULL; static void virAccessManagerDispose(void *obj); @@ -376,3 +378,23 @@ int virAccessManagerCheckStorageVol(virAccessManager *manager, return virAccessManagerSanitizeError(ret, driverName); } + +virDomainXMLOption *virAccessManagerGetXMLOption(void) +{ + return virAccessManagerXMLOption; +} + +void virAccessManagerSetXMLOption(virDomainXMLOption *xmlopt) +{ + virObjectUnref(virAccessManagerXMLOption); + + virAccessManagerXMLOption = virObjectRef(xmlopt); +} + +char *virAccessManagerGetXMLDesc(virDomainDef *domain) +{ + if (virAccessManagerGetXMLOption() != NULL) { + return virDomainDefFormat(domain, virAccessManagerGetXMLOption(), 0); + } + return NULL; +} diff --git a/src/access/viraccessmanager.h b/src/access/viraccessmanager.h index 2050ac9b85..3516f4f8e4 100644 --- a/src/access/viraccessmanager.h +++ b/src/access/viraccessmanager.h @@ -94,3 +94,6 @@ int virAccessManagerCheckStorageVol(virAccessManager *manager, virStoragePoolDef *pool, virStorageVolDef *vol, virAccessPermStorageVol perm); +virDomainXMLOption *virAccessManagerGetXMLOption(void); +void virAccessManagerSetXMLOption(virDomainXMLOption *xmlopt); +char *virAccessManagerGetXMLDesc(virDomainDef *domain); diff --git a/src/libvirt_private.syms b/src/libvirt_private.syms index 33b93cbd3e..8bcbc9c478 100644 --- a/src/libvirt_private.syms +++ b/src/libvirt_private.syms @@ -19,6 +19,7 @@ virAccessManagerGetDefault; virAccessManagerNew; virAccessManagerNewStack; virAccessManagerSetDefault; +virAccessManagerSetXMLOption; # access/viraccessperm.h diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index d2eddbd9ae..f9880f8ee3 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -105,6 +105,7 @@ #include "virdomaincheckpointobjlist.h" #include "virutil.h" #include "backup_conf.h" +#include "viraccessmanager.h" #define VIR_FROM_THIS VIR_FROM_QEMU @@ -835,6 +836,8 @@ qemuStateInitialize(bool privileged, defsecmodel))) goto error; + virAccessManagerSetXMLOption(qemu_driver->xmlopt); + qemu_driver->nbdkitCapsCache = qemuNbdkitCapsCacheNew(cfg->cacheDir); /* If hugetlbfs is present, then we need to create a sub-directory within -- 2.43.5

2 1