February 2024 - Devel - libvirt List Archives

[PATCH v2 00/17] qemu: Improve handling of architecture-specific defaults
by Andrea Bolognani 19 Feb '24

19 Feb '24

Changes from [v1]: * several patches have been pushed; * of the remaining changes, only the ones related to SCSI and USB controllers have been retained. I still intend to pursue the rest, but those two are where the real nasty stuff happens, so I'm focusing on them only for now; * improve the handling of USB controllers on s390x; * make all the code dealing with the legacy USB controller obsolete and get rid of it; * use out arguments to return models, making the new helpers fall in line with the usual libvirt API conventions. [v1] https://lists.libvirt.org/archives/list/devel@lists.libvirt.org/thread/G53M… Andrea Bolognani (17): tests: Add controller-scsi-default-unavailable qemu: Rename qemuDomainDefaultSCSIControllerModel() qemu: Move error reporting out of qemuDomainDefaultSCSIControllerModel() qemu: Improve qemuDomainDefaultSCSIControllerModel() qemu: Clean up qemuDomainDefaultSCSIControllerModel() qemu: Use virtio-scsi by default on RISC-V tests: Add usb-legacy-multiple tests: Add usb-legacy-device qemu: Always default to no USB controller on s390x qemu: Only use legacy USB controller if actually needed qemu: Validate USB controllers earlier qemu: Improve error message for USB controller validation qemu: Drop all code dealing with the legacy USB controller qemu: Add qemuDomainDefaultUSBControllerModel() qemu: Extend qemuDomainDefaultUSBControllerModel() qemu: Clean up qemuDomainDefaultUSBControllerModel() qemu: Use qemu-xhci by default on RISC-V src/qemu/qemu_command.c | 145 +-------- src/qemu/qemu_domain.c | 300 +++++++++++++----- src/qemu/qemu_domain.h | 6 +- src/qemu/qemu_hotplug.c | 21 +- src/qemu/qemu_validate.c | 69 +++- .../qemuhotplug-base-ccw-live+ccw-virtio.xml | 5 +- ...with-2-ccw-virtio+ccw-virtio-1-reverse.xml | 5 +- ...otplug-base-ccw-live-with-2-ccw-virtio.xml | 5 +- ...-with-ccw-virtio+ccw-virtio-2-explicit.xml | 5 +- ...-ccw-live-with-ccw-virtio+ccw-virtio-2.xml | 5 +- ...uhotplug-base-ccw-live-with-ccw-virtio.xml | 5 +- .../qemuhotplug-base-ccw-live.xml | 5 +- .../arm-vexpressa9-basic.aarch64-latest.args | 1 - .../arm-vexpressa9-nodevs.aarch64-latest.args | 1 - .../arm-vexpressa9-virtio.aarch64-latest.args | 1 - ...scsi-default-unavailable.x86_64-latest.err | 1 + .../controller-scsi-default-unavailable.xml | 15 + .../disk-arm-virtio-sd.aarch64-latest.args | 1 - ...ault-models.riscv64-latest.abi-update.args | 12 +- ...fault-models.riscv64-latest.abi-update.xml | 25 +- ...64-virt-default-models.riscv64-latest.args | 12 +- ...v64-virt-default-models.riscv64-latest.xml | 25 +- .../s390-usb-address.s390x-latest.xml | 6 +- .../sparc-minimal.sparc-latest.args | 1 - ...ault-unavailable-i440fx.x86_64-latest.args | 33 -- ...fault-unavailable-i440fx.x86_64-latest.err | 1 + ...fault-unavailable-i440fx.x86_64-latest.xml | 31 -- ...-default-unavailable-q35.x86_64-latest.err | 2 +- ...-default-unavailable-q35.x86_64-latest.xml | 33 -- ...ntroller-implicit-isapc.x86_64-latest.args | 1 - ...ler-nec-xhci-unavailable.x86_64-latest.xml | 33 -- .../usb-legacy-device.x86_64-latest.err | 1 + tests/qemuxmlconfdata/usb-legacy-device.xml | 15 + .../usb-legacy-multiple.x86_64-latest.err | 1 + tests/qemuxmlconfdata/usb-legacy-multiple.xml | 15 + tests/qemuxmlconftest.c | 28 +- 36 files changed, 418 insertions(+), 453 deletions(-) create mode 100644 tests/qemuxmlconfdata/controller-scsi-default-unavailable.x86_64-latest.err create mode 100644 tests/qemuxmlconfdata/controller-scsi-default-unavailable.xml delete mode 100644 tests/qemuxmlconfdata/usb-controller-default-unavailable-i440fx.x86_64-latest.args create mode 100644 tests/qemuxmlconfdata/usb-controller-default-unavailable-i440fx.x86_64-latest.err delete mode 100644 tests/qemuxmlconfdata/usb-controller-default-unavailable-i440fx.x86_64-latest.xml delete mode 100644 tests/qemuxmlconfdata/usb-controller-default-unavailable-q35.x86_64-latest.xml delete mode 100644 tests/qemuxmlconfdata/usb-controller-nec-xhci-unavailable.x86_64-latest.xml create mode 100644 tests/qemuxmlconfdata/usb-legacy-device.x86_64-latest.err create mode 100644 tests/qemuxmlconfdata/usb-legacy-device.xml create mode 100644 tests/qemuxmlconfdata/usb-legacy-multiple.x86_64-latest.err create mode 100644 tests/qemuxmlconfdata/usb-legacy-multiple.xml -- 2.43.0

4 43

[PATCH 0/3] syntax-check: Sync with gnulib, some tweaks
by Andrea Bolognani 19 Feb '24

19 Feb '24

One might see patch 3/3 and think hey, it would actually be nice if developers were getting warned that some checks haven't been executed because a tool is missing! And I completely agree. However, since meson needs to see an exit code of 77 to mark a test as skipped, and make return codes are in the range 0-2, making that happen would require taking the cppi/flake8/black checks out of the current scaffolding and wrap them in some other way. I have no intention to shave that specific yak today O:-) Andrea Bolognani (3): all: Don't use 'grep -q' syntax-check: Sync with gnulib syntax-check: Drop 'syntax-check' target build-aux/syntax-check.mk | 33 ++++++++------------------------- libvirt.spec.in | 2 +- tests/virsh-uriprecedence | 2 +- tests/virt-aa-helper-test | 2 +- 4 files changed, 11 insertions(+), 28 deletions(-) -- 2.43.0

3 5

[PATCH] Set stubDriverName from hostdev driver model attribute during pci device setup
by Laine Stump 19 Feb '24

19 Feb '24

commit v9.10.0-129-g8b93d78c83 (first appearing in libvirt-10.0.0) was supposed to allow forcing a PCI hostdev to be bound to a particular driver by adding <driver model='blah'/> to the XML for the device. Unfortunately, a single line was missed during the final changes to the patch prior to pushing, and the result was that the driver model could be set to *anything* and it would be accepted but just ignored. This patch adds the missing line, which will set the stubDriverName field of the virPCIDevice object from the hostdev object as the virPCIDevice is being created. This ends up being used by virPCIDeviceBindToStub() as the driver that it binds the device to. Fixes: 8b93d78c8325f1fba5db98848350f3db43f5e7d5 Signed-off-by: Laine Stump <laine(a)redhat.com> --- src/hypervisor/virhostdev.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/src/hypervisor/virhostdev.c b/src/hypervisor/virhostdev.c index 40f8a4bc2c..53327f4f4d 100644 --- a/src/hypervisor/virhostdev.c +++ b/src/hypervisor/virhostdev.c @@ -242,6 +242,8 @@ virHostdevGetPCIHostDevice(const virDomainHostdevDef *hostdev, return -1; virPCIDeviceSetManaged(actual, hostdev->managed); + virPCIDeviceSetStubDriverName(actual, pcisrc->driver.model); + if (pcisrc->driver.name == VIR_DEVICE_HOSTDEV_PCI_DRIVER_NAME_VFIO) { virPCIDeviceSetStubDriverType(actual, VIR_PCI_STUB_DRIVER_VFIO); -- 2.43.0

2 1

[PATCH 0/2] Introduce mshv Hypervisor type.
by Praveen K Paladugu 16 Feb '24

16 Feb '24

These patches introduce 'mshv' hypervisor type and check for the hypervisor device on host while starting ch guests. Praveen K Paladugu (2): conf: Introduce mshv hypervisor type ch: Check for hypervisor while starting guests src/ch/ch_conf.c | 2 ++ src/ch/ch_driver.c | 7 +++++++ src/ch/ch_process.c | 34 ++++++++++++++++++++++++++++++++++ src/conf/domain_conf.c | 1 + src/conf/domain_conf.h | 1 + src/qemu/qemu_command.c | 1 + 6 files changed, 46 insertions(+) -- 2.43.0

3 11

[PATCH 0/4] Protect secret values stored on disk with TPM
by Michal Privoznik 16 Feb '24

16 Feb '24

Secret values are stored effectively in plaintext on a disk and we rely on file perms to secure them. But with systemd-cred we can use system's TPM chip and encrypt them. Such secrets won't be transferable to another system by simply copying files stored on disk, but: a) that's not recommended way anyway, b) one can argue secrets shouldn't be migrated anyway. Future work consists of encrypting secret values even when stored in memory, as it's now possible to obtain secrets by dumping memory of virsecretd. Though, to dump a memory admin rights are required at which point users can just read values stored on disk (which is not true for ephemeral secrets). Michal Prívozník (4): virsecret: Introduce APIs to talk to systemd-cred conf: Introduce @tpm attribute to <secret/> virsecretobj: Encrypt/decrypt secrets using TPM NEWS: Document new virSecret TPM feature NEWS.rst | 6 + docs/formatsecret.rst | 8 +- src/conf/schemas/secret.rng | 5 + src/conf/secret_conf.c | 17 +++ src/conf/secret_conf.h | 2 + src/conf/virsecretobj.c | 32 ++++- src/libvirt_private.syms | 3 + src/secret/secret_driver.c | 7 + src/util/virsecret.c | 170 +++++++++++++++++++++++ src/util/virsecret.h | 10 ++ tests/secretxml2xmlin/usage-tpm-vtpm.xml | 7 + tests/secretxml2xmltest.c | 1 + 12 files changed, 263 insertions(+), 5 deletions(-) create mode 100644 tests/secretxml2xmlin/usage-tpm-vtpm.xml -- 2.43.0

3 8

[PATCH 0/2] docs: domain: Fixes
by Peter Krempa 16 Feb '24

16 Feb '24

Peter Krempa (2): docs: domain: Fix broken rST clauses docs: domain: Fix iothread to virtqueue mapping example docs/formatdomain.rst | 12 +++++------- 1 file changed, 5 insertions(+), 7 deletions(-) -- 2.43.0

3 4

[PATCH 0/2] Add locking option for disk
by Hiroki Narukawa 16 Feb '24

16 Feb '24

There is a case that Linux has a bug and unlocking does not work properly like this: https://lore.kernel.org/lkml/20230608084609.14245-1-zhangjiachen.jaycee@byt… Especiall in the situation that live migration source node has this kind of bug, destination must not locking, or otherwise VM stops when live migration runs. There commits add workaround for this kind of problems. Could you take a look? Especially following is points I could not make sure: In the first commit, I wrote explanation on file section in the document, but placed parser and formatter in virDomainStorageSourceParse and virDomainStorageSourceParse because locking can be common if added to other backends. Is it good to place here, or is there some better location to place? Hiroki Narukawa (2): conf: add locking option to disk source qemu_block: add locking option docs/formatdomain.rst | 5 ++++ src/conf/domain_conf.c | 8 +++++++ src/conf/schemas/domaincommon.rng | 5 ++++ src/conf/storage_source_conf.h | 3 +++ src/qemu/qemu_block.c | 7 ++++++ tests/qemublocktest.c | 1 + ...le-raw-aio_native-locking-off-srconly.json | 10 ++++++++ .../file-raw-aio_native-locking-off.json | 23 +++++++++++++++++++ .../file-raw-aio_native-locking-off.xml | 12 ++++++++++ 9 files changed, 74 insertions(+) create mode 100644 tests/qemublocktestdata/xml2json/file-raw-aio_native-locking-off-srconly.json create mode 100644 tests/qemublocktestdata/xml2json/file-raw-aio_native-locking-off.json create mode 100644 tests/qemublocktestdata/xml2json/file-raw-aio_native-locking-off.xml -- 2.25.1

2 6

[PATCH v2 00/11] nodedev state and update
by Boris Fiuczynski 15 Feb '24

15 Feb '24

The series add a dual state to the mdev node devices as these objects can be active and defined at the same time. These two states can become different. To be able to also introspect the persisted and transient nodedevs filtering is added. To be able to also dump the XML of an inactive state while the node device is active a new option is added. The last three patches add the capability to update a mdev node device. This can be done on the persisted configuration, on the active configuration or on both. To support this v1.3.0 of mdevctl is required. Changes since v1: * replaced spec file requirement for v1.3.0 of mdevctl by a dynamic support check and an unsupported message if not available. * removed persisted precheck in virsh * renamed persisted and persist into persistent * addressed all other review comments made on v1 * added NEWS Boris Fiuczynski (11): virmdev: prepare type and attributes for dual state node_device: refactor mdev attributes handling node_device: remove unnecessary checks in virNodeDeviceDefFormat nodedev: add an active config to mdev tools: add option inactive to nodedev-dumpxml nodedev: add persisted and transient filter on list tools: add switches persisted and transient to nodedev-list virsh: doc fix on nodedev-list api: add virNodeDeviceUpdate() nodedev: Implement virNodeDeviceUpdateXML virsh: add nodedev-update NEWS.rst | 7 + docs/manpages/virsh.rst | 36 +- include/libvirt/libvirt-nodedev.h | 31 ++ libvirt.spec.in | 1 + src/access/viraccessperm.c | 1 + src/access/viraccessperm.h | 6 + src/conf/node_device_conf.c | 76 +-- src/conf/node_device_conf.h | 14 +- src/conf/virnodedeviceobj.c | 50 ++ src/conf/virnodedeviceobj.h | 3 + src/driver-nodedev.h | 6 + src/libvirt-nodedev.c | 47 +- src/libvirt_private.syms | 1 + src/libvirt_public.syms | 5 + src/node_device/node_device_driver.c | 459 +++++++++++++++--- src/node_device/node_device_driver.h | 17 +- src/node_device/node_device_udev.c | 5 +- src/remote/remote_driver.c | 1 + src/remote/remote_protocol.x | 17 +- src/remote_protocol-structs | 6 + src/test/test_driver.c | 6 +- src/util/virmdev.h | 6 + ...60c_c60c_c60c_c60c_c60cc60cc60c_update.xml | 16 + tests/nodedevmdevctldata/mdevctl-modify.argv | 25 + tests/nodedevmdevctldata/mdevctl-modify.json | 1 + tests/nodedevmdevctltest.c | 94 +++- ...v_c60cc60c_c60c_c60c_c60c_c60cc60cc60c.xml | 14 + ...d_b7f0_4fea_b468_f1da537d301b_inactive.xml | 1 + ...v_c60cc60c_c60c_c60c_c60c_c60cc60cc60c.xml | 10 + ...c_c60c_c60c_c60c_c60cc60cc60c_inactive.xml | 9 + ...9_36ea_4111_8f0a_8c9a70e21366_inactive.xml | 1 + ...9_495e_4243_ad9f_beb3f14c23d9_inactive.xml | 1 + ...4_f554_4dc1_809d_b2a01e8e48ad_inactive.xml | 8 + ...6_1ca8_49ac_b176_871d16c13076_inactive.xml | 1 + tests/nodedevxml2xmltest.c | 59 ++- tools/virsh-nodedev.c | 132 ++++- 36 files changed, 1030 insertions(+), 143 deletions(-) create mode 100644 tests/nodedevmdevctldata/mdev_c60cc60c_c60c_c60c_c60c_c60cc60cc60c_update.xml create mode 100644 tests/nodedevmdevctldata/mdevctl-modify.argv create mode 100644 tests/nodedevmdevctldata/mdevctl-modify.json create mode 100644 tests/nodedevschemadata/mdev_c60cc60c_c60c_c60c_c60c_c60cc60cc60c.xml create mode 120000 tests/nodedevxml2xmlout/mdev_3627463d_b7f0_4fea_b468_f1da537d301b_inactive.xml create mode 100644 tests/nodedevxml2xmlout/mdev_c60cc60c_c60c_c60c_c60c_c60cc60cc60c.xml create mode 100644 tests/nodedevxml2xmlout/mdev_c60cc60c_c60c_c60c_c60c_c60cc60cc60c_inactive.xml create mode 120000 tests/nodedevxml2xmlout/mdev_d069d019_36ea_4111_8f0a_8c9a70e21366_inactive.xml create mode 120000 tests/nodedevxml2xmlout/mdev_d2441d39_495e_4243_ad9f_beb3f14c23d9_inactive.xml create mode 100644 tests/nodedevxml2xmlout/mdev_ee0b88c4_f554_4dc1_809d_b2a01e8e48ad_inactive.xml create mode 120000 tests/nodedevxml2xmlout/mdev_fedc4916_1ca8_49ac_b176_871d16c13076_inactive.xml -- 2.42.0

2 21

[PATCH 0/3] meson: Drop userfaultfd_sysctl option
by Andrea Bolognani 15 Feb '24

15 Feb '24

Recently introduced as part of [1]. We can skip the meson part for this one, and deal with everything in the spec file only. Test pipeline: https://gitlab.com/abologna/libvirt/-/pipelines/1176890275 [1] https://lists.libvirt.org/archives/list/devel@lists.libvirt.org/thread/DKS4… Andrea Bolognani (3): rpm: Always enable sysctl_config rpm: Always enable userfaultfd_sysctl meson: Drop userfaultfd_sysctl option libvirt.spec.in | 13 +++++-------- meson.build | 7 ------- meson_options.txt | 1 - src/qemu/meson.build | 2 +- 4 files changed, 6 insertions(+), 17 deletions(-) -- 2.43.0

2 6

[PUSHED PATCH] spec: Explicitly disable userfaultfd_sysctl for mingw
by Jiri Denemark 14 Feb '24

14 Feb '24

The %meson* macros pass --auto-features=enabled to enable all "auto" features, which means we have to explicitly disable them. Signed-off-by: Jiri Denemark <jdenemar(a)redhat.com> --- Notes: Pushed as a build breaker. libvirt.spec.in | 1 + 1 file changed, 1 insertion(+) diff --git a/libvirt.spec.in b/libvirt.spec.in index c5ba4b632f..49ce717e1b 100644 --- a/libvirt.spec.in +++ b/libvirt.spec.in @@ -1452,6 +1452,7 @@ export SOURCE_DATE_EPOCH=$(stat --printf='%Y' %{_specdir}/libvirt.spec) -Dstorage_vstorage=disabled \ -Dstorage_zfs=disabled \ -Dsysctl_config=disabled \ + -Duserfaultfd_sysctl=disabled \ -Dtests=disabled \ -Dudev=disabled \ -Dwireshark_dissector=disabled \ -- 2.43.0

4 3

[PATCH] virsh: completer for --type of commands find-storage-pool-sources and find-storage-pool-sources-as
by Adam Julis 14 Feb '24

14 Feb '24

Signed-off-by: Adam Julis <ajulis(a)redhat.com> --- tools/virsh-completer-pool.c | 11 +++++++++++ tools/virsh-completer-pool.h | 5 +++++ tools/virsh-pool.c | 2 ++ 3 files changed, 18 insertions(+) diff --git a/tools/virsh-completer-pool.c b/tools/virsh-completer-pool.c index 0600394411..1081e5c10c 100644 --- a/tools/virsh-completer-pool.c +++ b/tools/virsh-completer-pool.c @@ -66,6 +66,17 @@ virshStoragePoolNameCompleter(vshControl *ctl, } +char ** +virshStoragePoolTypeCompleter(vshControl *ctl G_GNUC_UNUSED, + const vshCmd *cmd G_GNUC_UNUSED, + unsigned int flags) +{ + virCheckFlags(0, NULL); + + return virshEnumComplete(VIR_STORAGE_POOL_LAST, virStoragePoolTypeToString); +} + + char ** virshPoolEventNameCompleter(vshControl *ctl G_GNUC_UNUSED, const vshCmd *cmd G_GNUC_UNUSED, diff --git a/tools/virsh-completer-pool.h b/tools/virsh-completer-pool.h index 059b6ba9e0..5a051d2aeb 100644 --- a/tools/virsh-completer-pool.h +++ b/tools/virsh-completer-pool.h @@ -32,6 +32,11 @@ virshPoolEventNameCompleter(vshControl *ctl, const vshCmd *cmd, unsigned int flags); +char ** +virshStoragePoolTypeCompleter(vshControl *ctl, + const vshCmd *cmd, + unsigned int completerflags); + char ** virshPoolTypeCompleter(vshControl *ctl, const vshCmd *cmd, diff --git a/tools/virsh-pool.c b/tools/virsh-pool.c index 36f00cf643..db5cb6306c 100644 --- a/tools/virsh-pool.c +++ b/tools/virsh-pool.c @@ -1414,6 +1414,7 @@ static const vshCmdOptDef opts_find_storage_pool_sources_as[] = { {.name = "type", .type = VSH_OT_DATA, .flags = VSH_OFLAG_REQ, + .completer = virshStoragePoolTypeCompleter, .help = N_("type of storage pool sources to find") }, {.name = "host", @@ -1501,6 +1502,7 @@ static const vshCmdOptDef opts_find_storage_pool_sources[] = { {.name = "type", .type = VSH_OT_DATA, .flags = VSH_OFLAG_REQ, + .completer = virshStoragePoolTypeCompleter, .help = N_("type of storage pool sources to discover") }, {.name = "srcSpec", -- 2.43.0

2 1

[libvirt PATCH v2 0/3] qemu: Add support for /dev/userfaultfd
by Jiri Denemark 14 Feb '24

14 Feb '24

Jiri Denemark (3): qemu: Add support for /dev/userfaultfd build: Add userfaultfd_sysctl build option spec: Disable with_userfaultfd_sysctl on Fedora and RHEL-9 libvirt.spec.in | 16 +++++++++++ meson.build | 8 ++++++ meson_options.txt | 1 + src/qemu/meson.build | 2 +- src/qemu/qemu.conf.in | 3 +- src/qemu/qemu_cgroup.c | 1 + src/qemu/qemu_process.c | 38 +++++++++++++++++++++++++ src/qemu/qemu_security.c | 45 ++++++++++++++++++++++++++++++ src/qemu/qemu_security.h | 5 ++++ src/qemu/test_libvirtd_qemu.aug.in | 1 + 10 files changed, 118 insertions(+), 2 deletions(-) -- 2.43.0

2 7

[PATCH 0/4] char: Minor fixes, and a tighter QAPI schema
by Markus Armbruster 13 Feb '24

13 Feb '24

Markus Armbruster (4): chardev/parallel: Don't close stdin on inappropriate device tests/unit/test-char: Fix qemu_socket(), make_udp_socket() check qapi/char: Make backend types properly conditional qapi/char: Deprecate backend type "memory" docs/about/deprecated.rst | 8 ++++++++ qapi/char.json | 28 +++++++++++++++++----------- include/qemu/osdep.h | 9 ++++++++- chardev/char-parallel.c | 7 +++++-- tests/unit/test-char.c | 25 +++++++++++++++++++++++-- chardev/meson.build | 4 +--- 6 files changed, 62 insertions(+), 19 deletions(-) -- 2.43.0

4 14

gitlab pages hosting for libvirt.org
by Peter Krempa 13 Feb '24

13 Feb '24

Hi, gitlab released the support for HTTP redirects to other domains: https://gitlab.com/gitlab-org/gitlab-pages/-/issues/601#note_1761877536 this means that our definitions of redirects which are needed to host our main web page via libvirt pages became valid: https://libvirt.gitlab.io/libvirt/_redirects https://gitlab.com/libvirt/libvirt/-/blob/master/.gitlab_pages_redirects?re… The redirects currently apply on the libvirt.gitlab.io/libvirt domain which test-hosts the page for now. If anyone wants to check the redirects for themselves please do so. Later today or tomorrow I'll push the already-reviewed patch: https://lists.libvirt.org/archives/list/devel@lists.libvirt.org/message/FFY… which will switch over the redirect rule prefixes for hosting from a custom domain (libvirt.org) thus checking them will not be possible until the hosting is switched over.

1 1

[PATCH 0/4] secret: Modernize XML parsing and formatting
by Michal Privoznik 13 Feb '24

13 Feb '24

Couple of things happening here: 1) add a missing case to our polkit checks 2) modernize XML parsing and formatting, so that the code is prepared for a new feature I'm working on. Michal Prívozník (4): viraccessdriverpolkit: Add missing vtpm case secret_conf: Simplify calling of virSecretDefParseUsage() virSecretDef: Convert 'usage_type' field to proper enum type secret_conf: Modernize XML parsing & formatting src/access/viraccessdriverpolkit.c | 17 ++++++ src/conf/secret_conf.c | 92 ++++++++++++++---------------- src/conf/secret_conf.h | 2 +- 3 files changed, 60 insertions(+), 51 deletions(-) -- 2.43.0

2 5

[PATCH 1/3] qemu: add a 'chain' parameter to nbdkit start/stop
by Jonathon Jongsma 12 Feb '24

12 Feb '24

This will allow us to start or stop nbdkit for just a single disk source or for every source in the backing chain. This will be used in following patches. Signed-off-by: Jonathon Jongsma <jjongsma(a)redhat.com> --- src/qemu/qemu_extdevice.c | 8 +++--- src/qemu/qemu_hotplug.c | 6 ++--- src/qemu/qemu_nbdkit.c | 51 ++++++++++++++++++++++++++++++--------- src/qemu/qemu_nbdkit.h | 6 +++-- 4 files changed, 51 insertions(+), 20 deletions(-) diff --git a/src/qemu/qemu_extdevice.c b/src/qemu/qemu_extdevice.c index 3cf3867056..ed5976d1f7 100644 --- a/src/qemu/qemu_extdevice.c +++ b/src/qemu/qemu_extdevice.c @@ -234,12 +234,12 @@ qemuExtDevicesStart(virQEMUDriver *driver, for (i = 0; i < def->ndisks; i++) { virDomainDiskDef *disk = def->disks[i]; - if (qemuNbdkitStartStorageSource(driver, vm, disk->src) < 0) + if (qemuNbdkitStartStorageSource(driver, vm, disk->src, true) < 0) return -1; } if (def->os.loader && def->os.loader->nvram) { - if (qemuNbdkitStartStorageSource(driver, vm, def->os.loader->nvram) < 0) + if (qemuNbdkitStartStorageSource(driver, vm, def->os.loader->nvram, true) < 0) return -1; } @@ -297,11 +297,11 @@ qemuExtDevicesStop(virQEMUDriver *driver, for (i = 0; i < def->ndisks; i++) { virDomainDiskDef *disk = def->disks[i]; - qemuNbdkitStopStorageSource(disk->src, vm); + qemuNbdkitStopStorageSource(disk->src, vm, true); } if (def->os.loader && def->os.loader->nvram) - qemuNbdkitStopStorageSource(def->os.loader->nvram, vm); + qemuNbdkitStopStorageSource(def->os.loader->nvram, vm, true); } diff --git a/src/qemu/qemu_hotplug.c b/src/qemu/qemu_hotplug.c index 31b00e05ca..e67673b762 100644 --- a/src/qemu/qemu_hotplug.c +++ b/src/qemu/qemu_hotplug.c @@ -1018,7 +1018,7 @@ qemuDomainAttachDeviceDiskLiveInternal(virQEMUDriver *driver, if (qemuHotplugAttachManagedPR(vm, disk->src, VIR_ASYNC_JOB_NONE) < 0) goto cleanup; - if (qemuNbdkitStartStorageSource(driver, vm, disk->src) < 0) + if (qemuNbdkitStartStorageSource(driver, vm, disk->src, true) < 0) goto cleanup; } @@ -1045,7 +1045,7 @@ qemuDomainAttachDeviceDiskLiveInternal(virQEMUDriver *driver, if (virStorageSourceChainHasManagedPR(disk->src)) ignore_value(qemuHotplugRemoveManagedPR(vm, VIR_ASYNC_JOB_NONE)); - qemuNbdkitStopStorageSource(disk->src, vm); + qemuNbdkitStopStorageSource(disk->src, vm, true); } qemuDomainSecretDiskDestroy(disk); qemuDomainCleanupStorageSourceFD(disk->src); @@ -4562,7 +4562,7 @@ qemuDomainRemoveDiskDevice(virQEMUDriver *driver, qemuHotplugRemoveManagedPR(vm, VIR_ASYNC_JOB_NONE) < 0) goto cleanup; - qemuNbdkitStopStorageSource(disk->src, vm); + qemuNbdkitStopStorageSource(disk->src, vm, true); if (disk->transient) { VIR_DEBUG("Removing transient overlay '%s' of disk '%s'", diff --git a/src/qemu/qemu_nbdkit.c b/src/qemu/qemu_nbdkit.c index 1c72b6fe6a..39f9c58a48 100644 --- a/src/qemu/qemu_nbdkit.c +++ b/src/qemu/qemu_nbdkit.c @@ -893,18 +893,34 @@ qemuNbdkitInitStorageSource(qemuNbdkitCaps *caps WITHOUT_NBDKIT_UNUSED, } +static int +qemuNbdkitStartStorageSourceOne(virQEMUDriver *driver, + virDomainObj *vm, + virStorageSource *src) +{ + qemuDomainStorageSourcePrivate *priv = QEMU_DOMAIN_STORAGE_SOURCE_PRIVATE(src); + + if (priv && priv->nbdkitProcess && + qemuNbdkitProcessStart(priv->nbdkitProcess, vm, driver) < 0) + return -1; + + return 0; +} + + int qemuNbdkitStartStorageSource(virQEMUDriver *driver, virDomainObj *vm, - virStorageSource *src) + virStorageSource *src, + bool chain) { virStorageSource *backing; - for (backing = src; backing != NULL; backing = backing->backingStore) { - qemuDomainStorageSourcePrivate *priv = QEMU_DOMAIN_STORAGE_SOURCE_PRIVATE(backing); + if (!chain) + return qemuNbdkitStartStorageSourceOne(driver, vm, src); - if (priv && priv->nbdkitProcess && - qemuNbdkitProcessStart(priv->nbdkitProcess, vm, driver) < 0) + for (backing = src; backing != NULL; backing = backing->backingStore) { + if (qemuNbdkitStartStorageSourceOne(driver, vm, backing) < 0) return -1; } @@ -912,18 +928,31 @@ qemuNbdkitStartStorageSource(virQEMUDriver *driver, } +static void +qemuNbdkitStopStorageSourceOne(virStorageSource *src, + virDomainObj *vm) +{ + qemuDomainStorageSourcePrivate *priv = QEMU_DOMAIN_STORAGE_SOURCE_PRIVATE(src); + + if (priv && priv->nbdkitProcess && + qemuNbdkitProcessStop(priv->nbdkitProcess, vm) < 0) + VIR_WARN("Unable to stop nbdkit for storage source '%s'", + qemuBlockStorageSourceGetStorageNodename(src)); +} + + void qemuNbdkitStopStorageSource(virStorageSource *src, - virDomainObj *vm) + virDomainObj *vm, + bool chain) { virStorageSource *backing; - for (backing = src; backing != NULL; backing = backing->backingStore) { - qemuDomainStorageSourcePrivate *priv = QEMU_DOMAIN_STORAGE_SOURCE_PRIVATE(backing); + if (! chain) + return qemuNbdkitStopStorageSourceOne(src, vm); - if (priv && priv->nbdkitProcess && - qemuNbdkitProcessStop(priv->nbdkitProcess, vm) < 0) - VIR_WARN("Unable to stop nbdkit for storage source '%s'", qemuBlockStorageSourceGetStorageNodename(src)); + for (backing = src; backing != NULL; backing = backing->backingStore) { + qemuNbdkitStopStorageSourceOne(backing, vm); } } diff --git a/src/qemu/qemu_nbdkit.h b/src/qemu/qemu_nbdkit.h index 853b2cca6f..637bf962a7 100644 --- a/src/qemu/qemu_nbdkit.h +++ b/src/qemu/qemu_nbdkit.h @@ -63,11 +63,13 @@ qemuNbdkitReconnectStorageSource(virStorageSource *source, int qemuNbdkitStartStorageSource(virQEMUDriver *driver, virDomainObj *vm, - virStorageSource *src); + virStorageSource *src, + bool chain); void qemuNbdkitStopStorageSource(virStorageSource *src, - virDomainObj *vm); + virDomainObj *vm, + bool chain); int qemuNbdkitStorageSourceManageProcess(virStorageSource *src, -- 2.43.0

2 9

[PATCH 0/5] tests: More USB and capabilities for s390x
by Andrea Bolognani 12 Feb '24

12 Feb '24

*** UNIVERSAL SERIAL BLURB *** https://gitlab.com/abologna/libvirt/-/commits/caps-s390x Andrea Bolognani (5): tests: Rename s390-usb-* tests: Minimize s390-usb-* tests: Add s390-usb-model tests: Update capabilities for QEMU 8.1.0 on s390x tests: Add capabilities for QEMU 8.2.0 on s390x tests/domaincapsdata/qemu_8.1.0.s390x.xml | 80 +- ...u_8.1.0.s390x.xml => qemu_8.2.0.s390x.xml} | 152 +- .../caps_8.1.0_s390x.replies | 5000 +++++---- .../qemucapabilitiesdata/caps_8.1.0_s390x.xml | 732 +- ...s390x.replies => caps_8.2.0_s390x.replies} | 9526 ++++++++++------- ...s_8.1.0_s390x.xml => caps_8.2.0_s390x.xml} | 2058 ++-- ...default-video-type-s390x.s390x-latest.args | 2 +- ...vfio-zpci-ccw-memballoon.s390x-latest.args | 2 +- .../launch-security-s390-pv.s390x-latest.args | 2 +- .../s390-allow-bogus-usb-controller.xml | 32 - .../s390-allow-bogus-usb-none.xml | 30 - ...t-cpu-kvm-ccw-virtio-4.2.s390x-latest.args | 2 +- .../s390-defaultconsole.s390x-latest.args | 2 +- .../s390-panic.s390x-latest.args | 2 +- ...rgs => s390-usb-address.s390x-latest.args} | 9 - ....xml => s390-usb-address.s390x-latest.xml} | 22 +- tests/qemuxmlconfdata/s390-usb-address.xml | 15 + ....args => s390-usb-model.s390x-latest.args} | 11 +- ...st.xml => s390-usb-model.s390x-latest.xml} | 26 +- tests/qemuxmlconfdata/s390-usb-model.xml | 13 + ...t.args => s390-usb-none.s390x-latest.args} | 12 +- ...est.xml => s390-usb-none.s390x-latest.xml} | 22 +- tests/qemuxmlconfdata/s390-usb-none.xml | 13 + tests/qemuxmlconftest.c | 5 +- 24 files changed, 10053 insertions(+), 7717 deletions(-) copy tests/domaincapsdata/{qemu_8.1.0.s390x.xml => qemu_8.2.0.s390x.xml} (79%) copy tests/qemucapabilitiesdata/{caps_8.1.0_s390x.replies => caps_8.2.0_s390x.replies} (93%) copy tests/qemucapabilitiesdata/{caps_8.1.0_s390x.xml => caps_8.2.0_s390x.xml} (96%) delete mode 100644 tests/qemuxmlconfdata/s390-allow-bogus-usb-controller.xml delete mode 100644 tests/qemuxmlconfdata/s390-allow-bogus-usb-none.xml rename tests/qemuxmlconfdata/{s390-allow-bogus-usb-none.s390x-latest.args => s390-usb-address.s390x-latest.args} (58%) copy tests/qemuxmlconfdata/{s390-allow-bogus-usb-controller.s390x-latest.xml => s390-usb-address.s390x-latest.xml} (55%) create mode 100644 tests/qemuxmlconfdata/s390-usb-address.xml rename tests/qemuxmlconfdata/{s390-allow-bogus-usb-controller.s390x-latest.args => s390-usb-model.s390x-latest.args} (58%) rename tests/qemuxmlconfdata/{s390-allow-bogus-usb-controller.s390x-latest.xml => s390-usb-model.s390x-latest.xml} (51%) create mode 100644 tests/qemuxmlconfdata/s390-usb-model.xml copy tests/qemuxmlconfdata/{s390-defaultconsole.s390x-latest.args => s390-usb-none.s390x-latest.args} (61%) rename tests/qemuxmlconfdata/{s390-allow-bogus-usb-none.s390x-latest.xml => s390-usb-none.s390x-latest.xml} (50%) create mode 100644 tests/qemuxmlconfdata/s390-usb-none.xml -- 2.43.0

2 6

[PULL 02/10] docs: mark CRIS support as deprecated
by Alex Bennée 09 Feb '24

09 Feb '24

This might be premature but while streamlining the avocado tests I realised the only tests we have are "check-tcg" ones. The ageing fedora-cris-cross image works well enough for developers but can't be used in CI as we need supported build platforms to build QEMU. Does this mean the writing is on the wall for this architecture? Cc: Rabin Vincent <rabinv(a)axis.com> Reviewed-by: Thomas Huth <thuth(a)redhat.com> Acked-by: Edgar E. Iglesias <edgar.iglesias(a)gmail.com> Reviewed-by: Thomas Huth <thuth(a)redhat.com> Signed-off-by: Alex Bennée <alex.bennee(a)linaro.org> Message-Id: <20240207163812.3231697-3-alex.bennee(a)linaro.org> diff --git a/docs/about/deprecated.rst b/docs/about/deprecated.rst index c7b95e6068e..7b0c59919e5 100644 --- a/docs/about/deprecated.rst +++ b/docs/about/deprecated.rst @@ -192,6 +192,13 @@ in the QEMU object model anymore. ``power5+``, ``power5+_v2.1``, an alias, but for consistency these will get removed in a future release, too. Use ``power5p_v2.1`` and ``power7p_v2.1`` instead. +CRIS CPU architecture (since 9.0) +''''''''''''''''''''''''''''''''' + +The CRIS architecture was pulled from Linux in 4.17 and the compiler +is no longer packaged in any distro making it harder to run the +``check-tcg`` tests. Unless we can improve the testing situation there +is a chance the code will bitrot without anyone noticing. System emulator machines ------------------------ -- 2.39.2

1 0

Question about libvirtxml and kubernetes/kubevirt conventions
by Alice Frosi 09 Feb '24

09 Feb '24

Hi everyone, I would like to bring up a possible issue with the libvirt-go-xml-module and request your advice. We are considering using libvirtxml [2] as an alternative of the custom types for the libvirt schema in KubeVirt [1]. Nonetheless, KubeVirt makes an effort to adhere to Kubernetes best guidelines. One thing I've noticed, though, is that libvirtml consistently uses uint or int rather than uint32 or int32. This is in opposition to the primitive types that Kubernetes [3] recommends; in fact, KubeVirt uses them in the schema.go [4]. Would it be possible to align the libvirtxml go library to the Kubernetes conventions? [1] https://github.com/kubevirt/kubevirt [2] https://github.com/kubevirt/kubevirt/issues/10844 [3] https://github.com/kubernetes/community/blob/master/contributors/devel/sig-… [4] https://github.com/kubevirt/kubevirt/blob/main/pkg/virt-launcher/virtwrap/a… Many thanks, Alice

3 5

[PATCH v2] docs: Improve documentation for dies and clusters
by Andrea Bolognani 08 Feb '24

08 Feb '24

I've seen examples in the wild of the cluster attribute having non-zero value on x86_64. This is obviously quite confusing, but it's the information that Linux exposes to userspace and we don't really have a way to tell apart a valid die/cluster ID from a dummy one. What ultimately matters is that the underlying assumptions about topology are respected, which they are: in the x86_64 cases that I have analyzed, for example, each "cluster" contained exactly one core, so any program that would use this information to influence guest topology decisions would be unaffected by the additional level showing up in the hierarchy. In an attempt to reduce confusion, remove any reference to any specific value for the attributes having any special meaning attached to it. In fact, since there are plans to make it possible to create guests with multiple CPU clusters on x86_64, rework the note into a more generic warning cautioning users that an attribute showing up here does not imply that the same attribute can be used when defining a guest CPU topology. Signed-off-by: Andrea Bolognani <abologna(a)redhat.com> --- docs/formatcaps.rst | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/docs/formatcaps.rst b/docs/formatcaps.rst index f37532296f..68477e639f 100644 --- a/docs/formatcaps.rst +++ b/docs/formatcaps.rst @@ -74,14 +74,14 @@ The ``<host/>`` element consists of the following child elements: ``die_id`` Identifier for the die the CPU is in. - Note that not all architectures support CPU dies: if the current - architecture doesn't, the value will be 0 for all CPUs. + Note that, even if this attribute is present, you might not be able to + define guests with multiple CPU dies. ``cluster_id`` Identifier for the cluster the CPU is in. - Note that not all architectures support CPU clusters: if the current - architecture doesn't, the value will be 0 for all CPUs. + Note that, even if this attribute is present, you might not be able to + define guests with multiple CPU clusters. ``core_id`` Identifier for the core the CPU is in. -- 2.43.0

2 1

[PATCH] docs: Improve documentation for dies and clusters
by Andrea Bolognani 08 Feb '24

08 Feb '24

I've seen examples in the wild of the cluster attribute having non-zero value on x86_64. This is obviously quite confusing, but it's the information that Linux exposes to userspace and we don't really have a way to tell apart a valid die/cluster ID from a dummy one. What ultimately matters is that the underlying assumptions about topology are respected, which they are: in the x86_64 cases that I have analyzed, for example, each "cluster" contained exactly one core, so any program that would use this information to influence guest topology decisions would be unaffected by the additional level showing up in the hierarchy. In an attempt to reduce confusion, document that the value for these attributes is not necessarily going to be zero. Signed-off-by: Andrea Bolognani <abologna(a)redhat.com> --- docs/formatcaps.rst | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) diff --git a/docs/formatcaps.rst b/docs/formatcaps.rst index f37532296f..c15d391b63 100644 --- a/docs/formatcaps.rst +++ b/docs/formatcaps.rst @@ -74,14 +74,18 @@ The ``<host/>`` element consists of the following child elements: ``die_id`` Identifier for the die the CPU is in. - Note that not all architectures support CPU dies: if the current - architecture doesn't, the value will be 0 for all CPUs. + Note that, while not all architectures support CPU dies, this attribute + will always be present in the capabilities XML. If the architecture + doesn't support them, the value will likely be 0 for all CPUs, but it + could also be some other arbitrary value. ``cluster_id`` Identifier for the cluster the CPU is in. - Note that not all architectures support CPU clusters: if the current - architecture doesn't, the value will be 0 for all CPUs. + Note that, while not all architectures support CPU clusters, this + attribute will always be present in the capabilities XML. If the + architecture doesn't support them, the value will likely be 0 for all + CPUs, but it could also be some other arbitrary value. ``core_id`` Identifier for the core the CPU is in. -- 2.43.0

2 10

[libvirt PATCH 0/2] ci: lcitool: sync dependencies with libvirt-ci repo
by Ján Tomko 08 Feb '24

08 Feb '24

Ján Tomko (2): ci: lcitool: sync dependencies with libvirt-ci repo ci: regenerate ci/buildenv/almalinux-8.sh | 1 + ci/buildenv/alpine-319.sh | 1 + ci/buildenv/alpine-edge.sh | 1 + ci/buildenv/centos-stream-8.sh | 1 + ci/buildenv/centos-stream-9.sh | 1 + ci/buildenv/debian-11-cross-aarch64.sh | 1 + ci/buildenv/debian-11-cross-armv6l.sh | 1 + ci/buildenv/debian-11-cross-armv7l.sh | 1 + ci/buildenv/debian-11-cross-i686.sh | 1 + ci/buildenv/debian-11-cross-mips64el.sh | 1 + ci/buildenv/debian-11-cross-mipsel.sh | 1 + ci/buildenv/debian-11-cross-ppc64le.sh | 1 + ci/buildenv/debian-11-cross-s390x.sh | 1 + ci/buildenv/debian-11.sh | 1 + ci/buildenv/debian-12-cross-aarch64.sh | 1 + ci/buildenv/debian-12-cross-armv6l.sh | 1 + ci/buildenv/debian-12-cross-armv7l.sh | 1 + ci/buildenv/debian-12-cross-i686.sh | 1 + ci/buildenv/debian-12-cross-mips64el.sh | 1 + ci/buildenv/debian-12-cross-mipsel.sh | 1 + ci/buildenv/debian-12-cross-ppc64le.sh | 1 + ci/buildenv/debian-12-cross-s390x.sh | 1 + ci/buildenv/debian-12.sh | 1 + ci/buildenv/debian-sid-cross-aarch64.sh | 1 + ci/buildenv/debian-sid-cross-armv6l.sh | 1 + ci/buildenv/debian-sid-cross-armv7l.sh | 1 + ci/buildenv/debian-sid-cross-i686.sh | 1 + ci/buildenv/debian-sid-cross-mips64el.sh | 1 + ci/buildenv/debian-sid-cross-ppc64le.sh | 1 + ci/buildenv/debian-sid-cross-s390x.sh | 1 + ci/buildenv/debian-sid.sh | 1 + ci/buildenv/fedora-38-cross-mingw32.sh | 1 + ci/buildenv/fedora-38-cross-mingw64.sh | 1 + ci/buildenv/fedora-38.sh | 1 + ci/buildenv/fedora-39.sh | 1 + ci/buildenv/fedora-rawhide-cross-mingw32.sh | 1 + ci/buildenv/fedora-rawhide-cross-mingw64.sh | 1 + ci/buildenv/fedora-rawhide.sh | 1 + ci/buildenv/opensuse-leap-15.sh | 1 + ci/buildenv/opensuse-tumbleweed.sh | 1 + ci/buildenv/ubuntu-2004.sh | 1 + ci/buildenv/ubuntu-2204.sh | 1 + ci/containers/almalinux-8.Dockerfile | 1 + ci/containers/alpine-319.Dockerfile | 1 + ci/containers/alpine-edge.Dockerfile | 1 + ci/containers/centos-stream-8.Dockerfile | 1 + ci/containers/centos-stream-9.Dockerfile | 1 + ci/containers/debian-11-cross-aarch64.Dockerfile | 3 ++- ci/containers/debian-11-cross-armv6l.Dockerfile | 3 ++- ci/containers/debian-11-cross-armv7l.Dockerfile | 3 ++- ci/containers/debian-11-cross-i686.Dockerfile | 3 ++- ci/containers/debian-11-cross-mips64el.Dockerfile | 3 ++- ci/containers/debian-11-cross-mipsel.Dockerfile | 3 ++- ci/containers/debian-11-cross-ppc64le.Dockerfile | 3 ++- ci/containers/debian-11-cross-s390x.Dockerfile | 3 ++- ci/containers/debian-11.Dockerfile | 1 + ci/containers/debian-12-cross-aarch64.Dockerfile | 3 ++- ci/containers/debian-12-cross-armv6l.Dockerfile | 3 ++- ci/containers/debian-12-cross-armv7l.Dockerfile | 3 ++- ci/containers/debian-12-cross-i686.Dockerfile | 3 ++- ci/containers/debian-12-cross-mips64el.Dockerfile | 3 ++- ci/containers/debian-12-cross-mipsel.Dockerfile | 3 ++- ci/containers/debian-12-cross-ppc64le.Dockerfile | 3 ++- ci/containers/debian-12-cross-s390x.Dockerfile | 3 ++- ci/containers/debian-12.Dockerfile | 1 + ci/containers/debian-sid-cross-aarch64.Dockerfile | 3 ++- ci/containers/debian-sid-cross-armv6l.Dockerfile | 3 ++- ci/containers/debian-sid-cross-armv7l.Dockerfile | 3 ++- ci/containers/debian-sid-cross-i686.Dockerfile | 3 ++- ci/containers/debian-sid-cross-mips64el.Dockerfile | 3 ++- ci/containers/debian-sid-cross-ppc64le.Dockerfile | 3 ++- ci/containers/debian-sid-cross-s390x.Dockerfile | 3 ++- ci/containers/debian-sid.Dockerfile | 1 + ci/containers/fedora-38-cross-mingw32.Dockerfile | 3 ++- ci/containers/fedora-38-cross-mingw64.Dockerfile | 3 ++- ci/containers/fedora-38.Dockerfile | 1 + ci/containers/fedora-39.Dockerfile | 1 + ci/containers/fedora-rawhide-cross-mingw32.Dockerfile | 3 ++- ci/containers/fedora-rawhide-cross-mingw64.Dockerfile | 3 ++- ci/containers/fedora-rawhide.Dockerfile | 1 + ci/containers/opensuse-leap-15.Dockerfile | 1 + ci/containers/opensuse-tumbleweed.Dockerfile | 1 + ci/containers/ubuntu-2004.Dockerfile | 1 + ci/containers/ubuntu-2204.Dockerfile | 1 + ci/lcitool/projects/libvirt+minimal.yml | 2 +- ci/lcitool/projects/libvirt.yml | 5 +++-- 86 files changed, 115 insertions(+), 30 deletions(-) -- 2.43.0

3 4

[libvirt PATCH] qemu: Add support for /dev/userfaultfd
by Jiri Denemark 08 Feb '24

08 Feb '24

/dev/userfaultfd device is preferred over userfaultfd syscall for post-copy migrations. Unless qemu driver is configured to disable mount namespace or to forbid access to /dev/userfaultfd in cgroup_device_acl, we will copy it to the limited /dev filesystem QEMU will have access to and label it appropriately. So in the default configuration post-copy migration will be allowed even without enabling vm.unprivileged_userfaultfd sysctl. Signed-off-by: Jiri Denemark <jdenemar(a)redhat.com> --- Notes: The question is what should we do with the src/qemu/postcopy-migration.sysctl file which is installed by libvirt.spec to /usr/lib/sysctl.d/60-qemu-postcopy-migration.conf by default. The file is now useless and should ideally be removed, but only when the host kernel is new enough to support /dev/userfaultfd src/qemu/qemu.conf.in | 3 +- src/qemu/qemu_cgroup.c | 1 + src/qemu/qemu_process.c | 38 +++++++++++++++++++++++++ src/qemu/qemu_security.c | 45 ++++++++++++++++++++++++++++++ src/qemu/qemu_security.h | 5 ++++ src/qemu/test_libvirtd_qemu.aug.in | 1 + 6 files changed, 92 insertions(+), 1 deletion(-) diff --git a/src/qemu/qemu.conf.in b/src/qemu/qemu.conf.in index 34025a02ef..f406df8749 100644 --- a/src/qemu/qemu.conf.in +++ b/src/qemu/qemu.conf.in @@ -565,7 +565,8 @@ #cgroup_device_acl = [ # "/dev/null", "/dev/full", "/dev/zero", # "/dev/random", "/dev/urandom", -# "/dev/ptmx", "/dev/kvm" +# "/dev/ptmx", "/dev/kvm", +# "/dev/userfaultfd" #] # # RDMA migration requires the following extra files to be added to the list: diff --git a/src/qemu/qemu_cgroup.c b/src/qemu/qemu_cgroup.c index 47402b3750..5a5ba763a0 100644 --- a/src/qemu/qemu_cgroup.c +++ b/src/qemu/qemu_cgroup.c @@ -41,6 +41,7 @@ const char *const defaultDeviceACL[] = { "/dev/null", "/dev/full", "/dev/zero", "/dev/random", "/dev/urandom", "/dev/ptmx", "/dev/kvm", + "/dev/userfaultfd", NULL, }; #define DEVICE_PTY_MAJOR 136 diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index 0a6c18a671..6e51d6586b 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -2882,6 +2882,40 @@ qemuProcessStartManagedPRDaemon(virDomainObj *vm) } +static int +qemuProcessAllowPostCopyMigration(virDomainObj *vm) +{ + qemuDomainObjPrivate *priv = vm->privateData; + virQEMUDriver *driver = priv->driver; + g_autoptr(virQEMUDriverConfig) cfg = virQEMUDriverGetConfig(driver); + const char *const *devices = (const char *const *) cfg->cgroupDeviceACL; + const char *uffd = "/dev/userfaultfd"; + int rc; + + if (!virFileExists(uffd)) { + VIR_DEBUG("%s is not supported by the host", uffd); + return 0; + } + + if (!devices) + devices = defaultDeviceACL; + + if (!g_strv_contains(devices, uffd)) { + VIR_DEBUG("%s is not allowed by device ACL", uffd); + return 0; + } + + VIR_DEBUG("Labeling %s in mount namespace", uffd); + if ((rc = qemuSecurityDomainSetMountNSPathLabel(driver, vm, uffd)) < 0) + return -1; + + if (rc == 1) + VIR_DEBUG("Mount namespace is not enabled, leaving %s as is", uffd); + + return 0; +} + + static int qemuProcessInitPasswords(virQEMUDriver *driver, virDomainObj *vm, @@ -7802,6 +7836,10 @@ qemuProcessLaunch(virConnectPtr conn, qemuProcessStartManagedPRDaemon(vm) < 0) goto cleanup; + VIR_DEBUG("Setting up permissions to allow post-copy migration"); + if (qemuProcessAllowPostCopyMigration(vm) < 0) + goto cleanup; + VIR_DEBUG("Setting domain security labels"); if (qemuSecuritySetAllLabel(driver, vm, diff --git a/src/qemu/qemu_security.c b/src/qemu/qemu_security.c index 8bcef14d08..4aaa863ae9 100644 --- a/src/qemu/qemu_security.c +++ b/src/qemu/qemu_security.c @@ -615,6 +615,51 @@ qemuSecurityDomainRestorePathLabel(virQEMUDriver *driver, } +/** + * qemuSecurityDomainSetMountNSPathLabel: + * + * Label given path in mount namespace. If mount namespace is not enabled, + * nothing is labeled at all. + * + * Because the label is only applied in mount namespace, there's no need to + * restore it. + * + * Returns 0 on success, + * 1 when mount namespace is not enabled, + * -1 on error. + */ +int +qemuSecurityDomainSetMountNSPathLabel(virQEMUDriver *driver, + virDomainObj *vm, + const char *path) +{ + int ret = -1; + + if (!qemuDomainNamespaceEnabled(vm, QEMU_DOMAIN_NS_MOUNT)) { + VIR_DEBUG("Not labeling '%s': mount namespace disabled for domain '%s'", + path, vm->def->name); + return 1; + } + + if (virSecurityManagerTransactionStart(driver->securityManager) < 0) + goto cleanup; + + if (virSecurityManagerDomainSetPathLabel(driver->securityManager, + vm->def, path, false) < 0) + goto cleanup; + + if (virSecurityManagerTransactionCommit(driver->securityManager, + vm->pid, false) < 0) + goto cleanup; + + ret = 0; + + cleanup: + virSecurityManagerTransactionAbort(driver->securityManager); + return ret; +} + + /** * qemuSecurityCommandRun: * @driver: the QEMU driver diff --git a/src/qemu/qemu_security.h b/src/qemu/qemu_security.h index 10f11771b4..41da33debc 100644 --- a/src/qemu/qemu_security.h +++ b/src/qemu/qemu_security.h @@ -110,6 +110,11 @@ int qemuSecurityDomainRestorePathLabel(virQEMUDriver *driver, virDomainObj *vm, const char *path); +int +qemuSecurityDomainSetMountNSPathLabel(virQEMUDriver *driver, + virDomainObj *vm, + const char *path); + int qemuSecurityCommandRun(virQEMUDriver *driver, virDomainObj *vm, virCommand *cmd, diff --git a/src/qemu/test_libvirtd_qemu.aug.in b/src/qemu/test_libvirtd_qemu.aug.in index e4cfde6cc7..b97e6de11e 100644 --- a/src/qemu/test_libvirtd_qemu.aug.in +++ b/src/qemu/test_libvirtd_qemu.aug.in @@ -67,6 +67,7 @@ module Test_libvirtd_qemu = { "5" = "/dev/urandom" } { "6" = "/dev/ptmx" } { "7" = "/dev/kvm" } + { "8" = "/dev/userfaultfd" } } { "save_image_format" = "raw" } { "dump_image_format" = "raw" } -- 2.43.0

2 1

[PATCH 00/11] webpage related fixes
by Peter Krempa 08 Feb '24

08 Feb '24

Patch 1 fixes the 'maven' redirect for gitlab pages by adding a /. Note that I didn't test that yet as it's hard to do with the redirects file, so this one will be tested in produciton. The rest of the series optimizes JS, CSS and XSLT assets for reuse in libvirt wiki by separating out stuff that is relevant only for libvirt.org. There will be a corresponding merge request to pull them into the wiki project. The resulting page can be browsed at: https://pipo.sk.gitlab.io/-/libvirt/-/jobs/6119138205/artifacts/website/ind… https://gitlab.com/pipo.sk/libvirt/-/pipelines/1168193309 Peter Krempa (11): gitlab_pages: Fix 'maven' redirect docs: page.xsl: Fix headerlinks with new docutils js: Separate the virt-tools blog planet loader code css: mobile: Split up libvirt.org specific styles from main template style css: Split out page templated styles to 'libvirt-template.css' css: Remove unused CSS for migration support matrix table css: Move styles for the generated API documents to libvirt-api.css css: Move generic styles to 'generic.css' docs: Move 'html including' from page.xsl to site.xsl docs: xsl: Properly propagate 'href_base' as XSL template parameter docs: page.xsl: Split up 'href_base' for assets and links .gitlab_pages_redirects | 2 +- docs/css/generic.css | 59 +++ docs/css/libvirt-api.css | 112 +++++ docs/css/libvirt-template.css | 236 +++++++++++ docs/css/libvirt.css | 417 ------------------- docs/css/main.css | 5 +- docs/css/meson.build | 5 +- docs/css/mobile-libvirt.css | 17 + docs/css/{mobile.css => mobile-template.css} | 15 +- docs/html/meson.build | 2 + docs/index.rst | 5 +- docs/js/main.js | 62 --- docs/js/meson.build | 1 + docs/js/virt-tools-blog-planet.js | 65 +++ docs/newapi.xsl | 3 +- docs/page.xsl | 37 +- docs/site.xsl | 17 +- 17 files changed, 535 insertions(+), 525 deletions(-) create mode 100644 docs/css/libvirt-api.css create mode 100644 docs/css/libvirt-template.css create mode 100644 docs/css/mobile-libvirt.css rename docs/css/{mobile.css => mobile-template.css} (84%) create mode 100644 docs/js/virt-tools-blog-planet.js -- 2.43.0

2 12

volume default permissions
by Memet Bilgin 07 Feb '24

07 Feb '24

Hello all, this is a feature request for something that was evidently discussed then dismissed a while back (c.f. https://lists.libvirt.org/archives/list/devel@lists.libvirt.org/thread/5YCW… ) The feature/bug in the above thread is a succinct description of the issue and I will not elaborate on it much, however this statement: > at this point changing the defaults could surprise some people and the current behavior hasn't generated many complaints, so best to just leave it as is. doesn't make much sense. I get that default behaviour should be backwards compatible. However, there is a legitimate need for this feature when using libvirt remotely (e.g via terraform over qemu:// or qemu+ssh://) as the users of the remote API don't have access to the uid's of the user and group and this means that as things are, a remote user can *only* reliably use root:root as the permission - else they need to use a hard coded magic number or have access to the host. Explicitly specifying -1 in the XML as the user and group ids should make the volume inherit the pool ids. This is completely backwards compatible because any such value being literally interpreted in the legacy code base would have resulted in an error anyways. What are your thoughts? Regards, Memet Bilgin

1 0