December 2025 - Devel - libvirt List Archives

[PATCH] virfiletest: Fix build on non-Linux
by Jiri Denemark 10 Dec '25

10 Dec '25

From: Jiri Denemark <jdenemar(a)redhat.com> Fixes: 121d179e068b584f62ea2c029d89a44e67c909c0 Signed-off-by: Jiri Denemark <jdenemar(a)redhat.com> --- tests/virfiletest.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/tests/virfiletest.c b/tests/virfiletest.c index ccd76a3fac..3f179a3f74 100644 --- a/tests/virfiletest.c +++ b/tests/virfiletest.c @@ -329,6 +329,7 @@ testFileIsSharedFSType(const void *opaque G_GNUC_UNUSED) } +#ifdef __linux__ static const char *shared_filesystems[] = { "/run/user/501/gvfs", "/nfs", @@ -338,6 +339,7 @@ static const char *shared_filesystems[] = { "/quobyte", NULL, }; +#endif static int testFileIsSharedFSOverride(const void *opaque G_GNUC_UNUSED) -- 2.52.0

1 0

[PATCH 0/4] util: Fix race when checking whether a path is on a shared FS
by Jiri Denemark 10 Dec '25

10 Dec '25

See 2/4 for details. Jiri Denemark (4): tests: Test virFileIsSharedFSOverride util: Fix race condition in virFileIsSharedFSType util: Fix race condition in virFileIsSharedFSOverride util: Rework virFileIsSharedFSOverride using virFileCheckParents src/util/virfile.c | 133 +++++++++++++++++++++++++++++--------------- tests/virfiletest.c | 69 +++++++++++++++++++++++ 2 files changed, 156 insertions(+), 46 deletions(-) -- 2.52.0

3 6

[PATCH 0/9] qemu: Fix few bugs in snapshot/save code (background snapshot saga part 1)
by Peter Krempa 09 Dec '25

09 Dec '25

While looking trhough the snapshot code I've noticed a few bugs: - memleak in cleanup code of migration to file (1/9) - wrong place where 'manual' external disk snapshot is taken (2/9) - bad interaction between 'manual' snapshot and the _QUIESCE flag (3/9) - constants for save image types not exposed (6/9) - use of wrong 'flags' in migration to file code (8/9) The series contains also few cleanups and improvements: - bash completion for save image type (7/9) - use of proper types for save image type (4,5/9) - avoid problems with 'flags' in qemuMigrationParamsForSave (9/9) Part 2 will focus on some code cleanups. Peter Krempa (9): qemuMigrationSrcToFile: Don't leak 'qemuFDPass' in cleanup path qemu: snapshot: Setup disks for manual snapshot only when the VM is actually paused qemuSnapshotPrepare: Prohibit 'manual' disk snapshot mode with VIR_DOMAIN_SNAPSHOT_CREATE_QUIESCE flag qemu: Use 'virQEMUSaveFormat' type everywhere except qemu_conf qemu: conf: Use proper type for (save|dump|snapshot)ImageFormat include: Create constants for save image format values virsh: Add completer for '--image-format' option of 'save' command qemuMigrationSrcToFile: Don't cross-contaminate 'flags' variable qemuMigrationParamsForSave: Don't take opaque 'flags' include/libvirt/libvirt-domain.h | 87 ++++++++++++++++++++++++++++++-- src/qemu/qemu_conf.c | 48 ++++++++++++------ src/qemu/qemu_conf.h | 7 +-- src/qemu/qemu_driver.c | 29 +++++++---- src/qemu/qemu_migration.c | 14 ++--- src/qemu/qemu_migration.h | 2 +- src/qemu/qemu_migration_params.c | 4 +- src/qemu/qemu_migration_params.h | 2 +- src/qemu/qemu_saveimage.c | 22 ++++---- src/qemu/qemu_saveimage.h | 24 ++------- src/qemu/qemu_saveimage_format.h | 25 +++++++++ src/qemu/qemu_snapshot.c | 14 +++-- tools/virsh-domain.c | 21 ++++++++ 13 files changed, 224 insertions(+), 75 deletions(-) create mode 100644 src/qemu/qemu_saveimage_format.h -- 2.52.0

2 10

[PATCH 0/2] qemu: Treat memory device source nodemask as strict NUMA policy
by Michal Privoznik 08 Dec '25

08 Dec '25

*** BLURB HERE *** Michal Prívozník (2): qemuxmlconftest: Introduce memory-hotplug-numa-preferred test case qemu: Treat memory device source nodemask as strict NUMA policy src/qemu/qemu_command.c | 6 ++ ...-hotplug-numa-preferred.x86_64-latest.args | 44 ++++++++++++ ...y-hotplug-numa-preferred.x86_64-latest.xml | 1 + .../memory-hotplug-numa-preferred.xml | 68 +++++++++++++++++++ tests/qemuxmlconftest.c | 1 + 5 files changed, 120 insertions(+) create mode 100644 tests/qemuxmlconfdata/memory-hotplug-numa-preferred.x86_64-latest.args create mode 120000 tests/qemuxmlconfdata/memory-hotplug-numa-preferred.x86_64-latest.xml create mode 100644 tests/qemuxmlconfdata/memory-hotplug-numa-preferred.xml -- 2.51.2

2 3

[PATCH] qemu: tpm: Account for possible migration without actually sharing storage
by Peter Krempa 08 Dec '25

08 Dec '25

From: Peter Krempa <pkrempa(a)redhat.com> The current logic in 'qemuTPMEmulatorBuildCommand' skips all setup if the *location* of the data is on what we'd consider shared storage. This means that if the location is not actually shared (e.g. it's shared betweeh some other hosts than the two doing the migration) and the path wasn't ever used (e.g. by migrating out) from the host where we're migrating into the complete setup of the location would be skipped even when it doesn't exist. Fix the logic by skipping only some of the setup steps so that 'qemuTPMEmulatorCreateStorage' can still create the storage if it doesn't exist. The rest of the code then needs to take the 'created' flag returned from 'qemuTPMEmulatorCreateStorage' into account. Fixes: 68103e9daf633b789428fedef56f816c92f6ee75 Signed-off-by: Peter Krempa <pkrempa(a)redhat.com> --- src/qemu/qemu_tpm.c | 26 +++++++++++++++++--------- 1 file changed, 17 insertions(+), 9 deletions(-) diff --git a/src/qemu/qemu_tpm.c b/src/qemu/qemu_tpm.c index 4c9445d72c..1ce6390fd5 100644 --- a/src/qemu/qemu_tpm.c +++ b/src/qemu/qemu_tpm.c @@ -158,6 +158,7 @@ qemuTPMEmulatorGetPid(const char *swtpmStateDir, /** * qemuTPMEmulatorCreateStorage: * @tpm: TPM definition for an emulator type + * @sharedStorageMigration: VM is being migrated with possibly shared storage * @created: a pointer to a bool that will be set to true if the * storage was created because it did not exist yet * @swtpm_user: The uid that needs to be able to access the directory @@ -169,6 +170,7 @@ qemuTPMEmulatorGetPid(const char *swtpmStateDir, */ static int qemuTPMEmulatorCreateStorage(virDomainTPMDef *tpm, + bool sharedStorageMigration, bool *created, uid_t swtpm_user, gid_t swtpm_group) @@ -187,8 +189,14 @@ qemuTPMEmulatorCreateStorage(virDomainTPMDef *tpm, *created = false; if (!virFileExists(source_path) || - virDirIsEmpty(source_path, true) > 0) + virDirIsEmpty(source_path, true) > 0) { *created = true; + } else { + /* If the location exists and is shared, we don't need to create it + * during migration */ + if (sharedStorageMigration) + return 0; + } if (virDirCreate(source_path, 0700, swtpm_user, swtpm_group, VIR_DIR_CREATE_ALLOW_EXIST) < 0) { @@ -809,16 +817,13 @@ qemuTPMEmulatorBuildCommand(virDomainTPMDef *tpm, run_setup = true; } - /* Do not create storage and run swtpm_setup on incoming migration over - * shared storage - */ on_shared_storage = virFileIsSharedFS(tpm->data.emulator.source_path, cfg->sharedFilesystems) == 1; - if (incomingMigration && on_shared_storage) - create_storage = false; if (create_storage) { - if (qemuTPMEmulatorCreateStorage(tpm, &created, + if (qemuTPMEmulatorCreateStorage(tpm, + incomingMigration && on_shared_storage, + &created, cfg->swtpm_user, cfg->swtpm_group) < 0) return NULL; run_setup = created; @@ -885,6 +890,9 @@ qemuTPMEmulatorBuildCommand(virDomainTPMDef *tpm, /* If swtpm supports it and the TPM state is stored on shared storage, * start swtpm with --migration release-lock-outgoing so it can migrate * across shared storage if needed. + * + * Note that if 'created' is true, the location didn't exist so the storage + * is not actually shared. */ QEMU_DOMAIN_TPM_PRIVATE(tpm)->swtpm.can_migrate_shared_storage = false; if (on_shared_storage && @@ -892,13 +900,13 @@ qemuTPMEmulatorBuildCommand(virDomainTPMDef *tpm, virCommandAddArg(cmd, "--migration"); virCommandAddArgFormat(cmd, "release-lock-outgoing%s", - incomingMigration ? ",incoming": ""); + incomingMigration && !created ? ",incoming": ""); QEMU_DOMAIN_TPM_PRIVATE(tpm)->swtpm.can_migrate_shared_storage = true; } else { /* Report an error if there's an incoming migration across shared * storage and swtpm does not support the --migration option. */ - if (incomingMigration && on_shared_storage) { + if (incomingMigration && on_shared_storage && !created) { virReportError(VIR_ERR_ARGUMENT_UNSUPPORTED, _("%1$s (on destination side) does not support the --migration option needed for migration with shared storage"), swtpm); -- 2.52.0

2 3

[PATCH v2] NEWS: Mention the passt hostname and fqdn attributes
by Han Han 08 Dec '25

08 Dec '25

Signed-off-by: Han Han <hhan(a)redhat.com> --- V1: https://lists.libvirt.org/archives/list/devel@lists.libvirt.org/thread/NT3X… NEWS.rst | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/NEWS.rst b/NEWS.rst index 01efe82c97..28641ec1a1 100644 --- a/NEWS.rst +++ b/NEWS.rst @@ -217,6 +217,11 @@ v11.8.0 (2025-10-01) On the XML side, the existing ``<acpi>`` element has been extended with a ``nodeset`` attribute to specify the NUMA node affinity of a PCI device. + * qemu: Add support for hostname and FQDN configration of passt backend + + The attributes ``hostname`` and ``fqdn`` for passt backend configure + the guest interface with hostname and FQDN. + * **Improvements** * ch: Events emitting -- 2.52.0

2 1

[PATCH 0/1] qemu: Use pci_bus to identify multi-smmuv3 model
by Nathan Chen 05 Dec '25

05 Dec '25

The previous qemu virt-machine smmuv3 model and the new arm-smmuv3 model cannot both be specified at the same time. When launching a Libvirt VM with a single arm-smmuv3 model denoted by the <pciBus> attribute, qemuBuildMachineCommandLine() still appends the iommu=smmuv3 -machine parameter in addition to the arm-smmuv3 model passed to -device, preventing the VM from booting. Use the presence of a non-negative pci_bus to identify the multi-smmuv3 IOMMU model, instead of the niommus attribute. This allows for specifying a single arm-smmuv3 on the qemu command line, instead of both the virt-machine smmuv3 and arm-smmuv3 being specified at the same time. This patch is available on Github [0]. [0] https://github.com/NathanChenNVIDIA/libvirt/commits/single-smmuv3-12-25/ Signed-off-by: Nathan Chen <nathanc(a)nvidia.com> Nathan Chen (1): qemu: Use pci_bus to identify multi-smmuv3 model src/qemu/qemu_command.c | 2 +- src/qemu/qemu_postparse.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) -- 2.43.0

2 5

[PULL 14/14] aspeed: Deprecate the fby35 machine
by Alex Bennée 05 Dec '25

05 Dec '25

From: Cédric Le Goater <clg(a)redhat.com> There are no functional tests for the 'fby35' machine which makes harder to determine when something becomes deprecated or unused. The 'fby35' machine was originally added as an example of a multi-SoC system, with the expectation the models would evolve over time in an heterogeneous system. This hasn't happened and no public firmware is available to boot it. It can be replaced by the 'ast2700fc', another multi-SoC machine based on the newer AST2700 SoCs which are excepted to receive better support in the future. Cc: Peter Delevoryas <peter(a)pjd.dev> Signed-off-by: Cédric Le Goater <clg(a)redhat.com> Message-ID: <20251126102424.927527-1-clg(a)redhat.com> Signed-off-by: Alex Bennée <alex.bennee(a)linaro.org> diff --git a/docs/about/deprecated.rst b/docs/about/deprecated.rst index 9386cffba26..ac31a2bce42 100644 --- a/docs/about/deprecated.rst +++ b/docs/about/deprecated.rst @@ -308,13 +308,24 @@ I2C devices connected to the board can be defined via the QEMU command line. Arm ``fp5280g2-bmc`` machine (since 10.2) -''''''''''''''''''''''''''''''''''''''''''' +''''''''''''''''''''''''''''''''''''''''' The ``fp5280g2-bmc`` machine does not rely on any specific device models, it can be replaced by the ``ast2500-evb`` machine using the ``fmc-model`` option to specify the flash type. The I2C devices connected to the board can be defined via the QEMU command line. +Arm ``fby35`` machine (since 10.2) +'''''''''''''''''''''''''''''''''' + +The ``fby35`` machine was originally added as an example of a +multi-SoC system, with the expectation the models would evolve over +time in an heterogeneous system. This hasn't happened and no public +firmware is available to boot it. It can be replaced by the +``ast2700fc``, another multi-SoC machine based on the newer AST2700 +SoCs which are excepted to receive better support in the future. + + RISC-V default machine option (since 10.0) '''''''''''''''''''''''''''''''''''''''''' diff --git a/hw/arm/fby35.c b/hw/arm/fby35.c index 70d3a99d79e..e61a6438fef 100644 --- a/hw/arm/fby35.c +++ b/hw/arm/fby35.c @@ -175,6 +175,7 @@ static void fby35_class_init(ObjectClass *oc, const void *data) MachineClass *mc = MACHINE_CLASS(oc); mc->desc = "Meta Platforms fby35"; + mc->deprecation_reason = "For a multi-soc machine, use 'ast2700fc' instead"; mc->init = fby35_init; mc->no_floppy = 1; mc->no_cdrom = 1; -- 2.47.3

1 0

[PATCH] ci: refresh with 'lcitool manifest'
by Michal Privoznik 03 Dec '25

03 Dec '25

From: Michal Privoznik <mprivozn(a)redhat.com> Update FreeBSD to their new minor versions. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- Green pipeline (although it doesn't show cirrus jobs): https://gitlab.com/MichalPrivoznik/libvirt/-/pipelines/2194225996 There is also FreeBSD 15.0 which was released yesterday. But there are some issues with its availability in cirrus [1] so I'm not updating the majo version, for now. 1: https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=291354 ci/gitlab/builds.yml | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/ci/gitlab/builds.yml b/ci/gitlab/builds.yml index a98b0d9c46..fe1468ff81 100644 --- a/ci/gitlab/builds.yml +++ b/ci/gitlab/builds.yml @@ -515,7 +515,7 @@ x86_64-freebsd-13: allow_failure: exit_codes: 3 variables: - CIRRUS_VM_IMAGE_NAME: freebsd-13-4 + CIRRUS_VM_IMAGE_NAME: freebsd-13-5 CIRRUS_VM_IMAGE_SELECTOR: image_family CIRRUS_VM_INSTANCE_TYPE: freebsd_instance INSTALL_COMMAND: pkg install -y @@ -530,7 +530,7 @@ x86_64-freebsd-14: allow_failure: exit_codes: 3 variables: - CIRRUS_VM_IMAGE_NAME: freebsd-14-2 + CIRRUS_VM_IMAGE_NAME: freebsd-14-3 CIRRUS_VM_IMAGE_SELECTOR: image_family CIRRUS_VM_INSTANCE_TYPE: freebsd_instance INSTALL_COMMAND: pkg install -y -- 2.51.2

2 1

[PATCH] NEWS: Mention the passt hostname and fqdn attributes
by Han Han 03 Dec '25

03 Dec '25

Signed-off-by: Han Han <hhan(a)redhat.com> --- NEWS.rst | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/NEWS.rst b/NEWS.rst index 91ed53ba7e..1b2b8778b1 100644 --- a/NEWS.rst +++ b/NEWS.rst @@ -205,6 +205,12 @@ v11.8.0 (2025-10-01) On the XML side, the existing ``<acpi>`` element has been extended with a ``nodeset`` attribute to specify the NUMA node affinity of a PCI device. + * qemu: Add support for hostname and FQDN configration of passt backend + + Implement passt ``--hostname`` and ``--hostname`` options to libvirt via the + interface attributes ``hostname`` and ``fqdn`` of passt backend. They configure + the guest interface with hostname and FQDN. + * **Improvements** * ch: Events emitting -- 2.52.0

2 2