May 2026 - Devel - libvirt List Archives

[PATCH 00/10] qemu: Fix CPU feature reporting on AMD CPUs
by Jiri Denemark 28 May '26

28 May '26

See the last patch for detailed description. Jiri Denemark (10): qemu_capabilities: Split virQEMUCapsFillDomainCPUCaps qemu: Move domain caps flags handling to virQEMUCapsFillDomainCPUHostModel qemu_capabilities: Use g_autoptr in virQEMUCapsInitHostCPUModel qemu_capabilities: Split conditions in virQEMUCapsInitHostCPUModel qemu_capabilities: Cache expanded CPU util: Publish and mock virHostCPUGetMSRFromKVM cpu_x86: Introduce virCPUx86DataAddMSR cpu: Introduce virCPUUpdateFeatures Clarify documentation of VIR_CONNECT_GET_DOMAIN_CAPABILITIES_EXPAND_CPU_FEATURES qemu_capabilities: Fix domain capabilities on AMD CPUs docs/manpages/virsh.rst | 4 +- include/libvirt/libvirt-domain.h | 2 +- src/cpu/cpu.c | 33 +++++ src/cpu/cpu.h | 12 ++ src/cpu/cpu_x86.c | 54 ++++++-- src/cpu/cpu_x86.h | 4 + src/libvirt-domain.c | 7 +- src/libvirt_private.syms | 3 + src/qemu/qemu_capabilities.c | 223 ++++++++++++++++++++++--------- src/qemu/qemu_capabilities.h | 7 +- src/qemu/qemu_conf.c | 6 +- src/qemu/qemu_conf.h | 3 +- src/qemu/qemu_driver.c | 15 +-- src/util/virhostcpu.c | 16 ++- src/util/virhostcpu.h | 3 + tests/domaincapstest.c | 2 +- tests/qemucpumock.c | 22 +++ tools/virsh-host.c | 2 +- 18 files changed, 319 insertions(+), 99 deletions(-) -- 2.54.0

3 12

[PATCH] qemu_nbdkit: Fix format when printing time_t values
by Michal Privoznik 28 May '26

28 May '26

From: Michal Privoznik <mprivozn(a)redhat.com> The time_t type can be 32bit or 64bit signed integer. There are systems where it's defined as long, or long long (32bit systems usually). Therefore, using just 'l' length modifier is not good enough. Also, using 'u' conversion specifier is also wrong (though, values stored in qemuNbdkitCaps struct reflect mtime of some files, so there won't be a negative value). Anyway, do what we already do for virQEMUCaps - use '%lld' printf format and typecast to long long. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- Spotted this on the following pipeline: https://gitlab.com/MichalPrivoznik/libvirt/-/jobs/14580340190 in which I was testing moving our CI from Debian 12 to 13. Apparently, in Debian they decided to -D_TIME_BITS=64 for everything [1] and only when compiling on a 32bit arch it's when time_t expands to 'long long'. 1: https://wiki.debian.org/ReleaseGoals/64bit-time src/qemu/qemu_nbdkit.c | 22 +++++++++++----------- 1 file changed, 11 insertions(+), 11 deletions(-) diff --git a/src/qemu/qemu_nbdkit.c b/src/qemu/qemu_nbdkit.c index 542a6b1f44..6918fc6631 100644 --- a/src/qemu/qemu_nbdkit.c +++ b/src/qemu/qemu_nbdkit.c @@ -544,18 +544,18 @@ qemuNbdkitCapsFormatCache(qemuNbdkitCaps *nbdkitCaps) virBufferEscapeString(&buf, "<path>%s</path>\n", nbdkitCaps->path); - virBufferAsprintf(&buf, "<nbdkitctime>%lu</nbdkitctime>\n", - nbdkitCaps->ctime); + virBufferAsprintf(&buf, "<nbdkitctime>%lld</nbdkitctime>\n", + (long long)nbdkitCaps->ctime); virBufferEscapeString(&buf, "<plugindir>%s</plugindir>\n", nbdkitCaps->pluginDir); - virBufferAsprintf(&buf, "<plugindirmtime>%lu</plugindirmtime>\n", - nbdkitCaps->pluginDirMtime); + virBufferAsprintf(&buf, "<plugindirmtime>%lld</plugindirmtime>\n", + (long long)nbdkitCaps->pluginDirMtime); virBufferEscapeString(&buf, "<filterdir>%s</filterdir>\n", nbdkitCaps->filterDir); - virBufferAsprintf(&buf, "<filterdirmtime>%lu</filterdirmtime>\n", - nbdkitCaps->filterDirMtime); - virBufferAsprintf(&buf, "<selfctime>%lu</selfctime>\n", - nbdkitCaps->libvirtCtime); + virBufferAsprintf(&buf, "<filterdirmtime>%lld</filterdirmtime>\n", + (long long)nbdkitCaps->filterDirMtime); + virBufferAsprintf(&buf, "<selfctime>%lld</selfctime>\n", + (long long)nbdkitCaps->libvirtCtime); virBufferAsprintf(&buf, "<selfvers>%u</selfvers>\n", nbdkitCaps->libvirtVersion); @@ -593,10 +593,10 @@ virNbdkitCapsSaveFile(void *data, return -1; } - VIR_DEBUG("Saved caps '%s' for '%s' with (%lu, %lu)", + VIR_DEBUG("Saved caps '%s' for '%s' with (%lld, %lld)", filename, nbdkitCaps->path, - nbdkitCaps->ctime, - nbdkitCaps->libvirtCtime); + (long long)nbdkitCaps->ctime, + (long long)nbdkitCaps->libvirtCtime); return 0; } -- 2.53.0

2 1

Re: something wrong with the git log in libvirt recently
by Peter Krempa 28 May '26

28 May '26

(you've posted this to the old list address; I've added it to CC. Fix your contact list to point to devel(a)lists.libvirt.org) On Wed, May 20, 2026 at 15:24:52 +0200, Denis V. Lunev wrote: > Hi! > Hi, it's an artifact of people with mail domains with DMARC posting to the list. New mailman can't keep it's fingers from messing with the message, which breaks signatures, thus the sender needs to be rewritten. It's from people not setting git properly: https://www.libvirt.org/submitting-patches.html With the setting git will format an extra From which fixes the header messed up by mailman. Reviewers should notify senders, but stuff slips through. > Statistics will be rotten. I don't think this matters too much. > > I think mailmap need to be involved. I don't think mailmap can fix it. > > Den > > commit 19e1dcf860fab060e19ab97005a4f58faf016093 > Author: Akash Kulhalli via Devel <devel(a)lists.libvirt.org> > Date: Wed Apr 29 17:43:29 2026 +0530 > > API/qemu: add async unplug flag to virDomainSetVcpu > > Add VIR_DOMAIN_SETVCPU_ASYNC_UNPLUG for virDomainSetVcpu(). > > Define a dedicated virDomainSetVcpuBehaviour flag type and wire the > new flag through the QEMU driver. As with setvcpus async unplug, > success indicates request acceptance while final completion is > reported by the vcpu-removed event. > > Update the API documentation and add virsh support for the async path to > the setvcpu subcommand. > > Signed-off-by: Akash Kulhalli <akash.kulhalli(a)oracle.com> > Reviewed-by: Peter Krempa <pkrempa(a)redhat.com> I'm sorry I didn't notice this one.

5 8

[PATCH] docs: drvbhyve: reorganize sections about resource limits
by Roman Bogorodskiy 28 May '26

28 May '26

Sections about resource limits are currently placed across the file in between various device examples. To make it easier to follow, group them into a single section. This also allows to give an introduction on the rctl(8) framework once instead of repeating it for every resource type. Also, document the memory limitation support added in libvirt 12.4.0. Signed-off-by: Roman Bogorodskiy <bogorodskiy(a)gmail.com> --- docs/drvbhyve.rst | 75 +++++++++++++++++++++++++++++++---------------- 1 file changed, 50 insertions(+), 25 deletions(-) diff --git a/docs/drvbhyve.rst b/docs/drvbhyve.rst index 79418537df..2e5e361115 100644 --- a/docs/drvbhyve.rst +++ b/docs/drvbhyve.rst @@ -773,26 +773,6 @@ Example: Please refer to ``cam(4)``, ``ctl(4)``, and ``ctld(8)`` manual pages for more details on CAM and CTL. -vCPU pinning -~~~~~~~~~~~~ - -:since:`Since 12.1.0`, it is possible to pin domain vCPUs -to the specific host CPUs. - -Example: - -:: - - <domain type='bhyve'> - ... - <vcpu>2</vcpu> - <cputune> - <vcpupin vcpu="0" cpuset="1-4,^2"/> - <vcpupin vcpu="1" cpuset="0,4"/> - </cputune> - ... - </domain> - NUMA domains configuration ~~~~~~~~~~~~~~~~~~~~~~~~~~ @@ -835,13 +815,43 @@ Example:: Bhyve supports up to 16 ports per console. -Block I/O Tuning -~~~~~~~~~~~~~~~~ -:since:`Since 12.3.0`, it is possible to tune domain I/O. -This works on top of the +Resource tuning and limiting +---------------------------- +The libvirt bhyve driver supports tuning and limiting resources such as +CPU, memory, and I/O. + +For managing the memory and I/O limits, the bhyve driver uses the `rctl(4) <https://man.freebsd.org/cgi/man.cgi?query=rctl&sektion=4>`__ framework. -Sample configuration:: + +As of `FreeBSD 15.0-RELEASE`, rctl is not enabled by default. +Please refer to the manual page above and the +`FreeBSD Handbook <https://docs.freebsd.org/en/books/handbook/book/#security-rctl>`__ +on how to enable it. + +CPU tuning does not require any additional configuration. + +vCPU pinning +~~~~~~~~~~~~ +Pinning domain vCPUs to the specific host CPUs is supported :since:`since 12.1.0`. + +Example: + +:: + + <domain type='bhyve'> + ... + <vcpu>2</vcpu> + <cputune> + <vcpupin vcpu="0" cpuset="1-4,^2"/> + <vcpupin vcpu="1" cpuset="0,4"/> + </cputune> + ... + </domain> + +Block I/O tuning +~~~~~~~~~~~~~~~~ +Block I/O tuning is supported :since:`since 12.3.0`. Sample configuration:: <blkiotune> <device> @@ -857,6 +867,21 @@ The ``*`` path here means that the limits are applied to the domain as a whole. Currently, it is not possible to apply limits to the individual devices of the domain. +Memory limitation +~~~~~~~~~~~~~~~~~ +Setting a memory hard limit for a domain is supported :since:`since 12.4.0`. +Example:: + + <memtune> + <hard_limit unit='M'>512</hard_limit> + </memtune> + +Please refer to the `format domain <formatdomain.html#memory-tuning>`__ page +for considerations on setting ``hard_limit``. + +Currently, other memory tuning options (``soft_limit``, ``swap_hard_limit``, +and ``min_guarantee``) are not supported. + Guest-specific considerations ----------------------------- -- 2.52.0

2 1

[PATCH] NEWS: document bhyve changes for 12.4.0
by Roman Bogorodskiy 28 May '26

28 May '26

Signed-off-by: Roman Bogorodskiy <bogorodskiy(a)gmail.com> --- NEWS.rst | 32 ++++++++++++++++++++++++++++++++ 1 file changed, 32 insertions(+) diff --git a/NEWS.rst b/NEWS.rst index 8d1b927c41..4d9e62f9f5 100644 --- a/NEWS.rst +++ b/NEWS.rst @@ -23,6 +23,38 @@ v12.4.0 (unreleased) This allows to monitor per-VM energy consumption on supported platforms. Implemented via ``energytune`` element in ``cputune`` . + * bhyve: Add memory tuning support + + The bhyve driver now allows setting a domain's memory hard limit using the + following syntax:: + + <memtune> + <hard_limit unit='G'>1</hard_limit> + </memtune> + + It also implements the ``virDomainGetMemoryParameters()`` and + ``virDomainSetMemoryParameters()`` APIs for controlling + the memory hard limit of a running domain. + + * bhyve: Add virtio-console device support + + Domain XML can now use ``virtio-console`` devices. Among other things, + these devices can be used to configure a QEMU Guest Agent:: + + <devices> + <channel type='unix'> + <source mode='bind' path='/var/run/libvirt/bhyve/bhyve.agent'/> + <target type='virtio' name='org.qemu.guest_agent.0'/> + <address type='virtio-serial' controller='0' bus='0' port='1'/> + </channel> + </devices> + + Additionally, the ``virDomainQemuAgentCommand()`` API which allows + running arbitrary guest agent commands is now supported. + + The ``virDomainGetHostname()`` API, used to query the hostname of a guest, + now supports using the guest agent as a source of information. + * **Improvements** * **Bug fixes** -- 2.52.0

2 1

[PATCH] NEWS: Document features/improvements/bug fixes I've participated in
by Michal Privoznik 28 May '26

28 May '26

From: Michal Privoznik <mprivozn(a)redhat.com> There are some features/improvements/bug fixes I've either contributed or reviewed/merged. Document them for upcoming release. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- NEWS.rst | 33 +++++++++++++++++++++++++++++++++ 1 file changed, 33 insertions(+) diff --git a/NEWS.rst b/NEWS.rst index 8d1b927c41..e3b262ef3d 100644 --- a/NEWS.rst +++ b/NEWS.rst @@ -15,6 +15,10 @@ v12.4.0 (unreleased) * **Removed features** + * qemu: Bump minimum version to QEMU-7.2 + + The minimal required version of QEMU was bumped to 7.2.0. + * **New features** * resctrl: Add energy monitoring via resctrl's PERF_PKG_MON @@ -23,8 +27,25 @@ v12.4.0 (unreleased) This allows to monitor per-VM energy consumption on supported platforms. Implemented via ``energytune`` element in ``cputune`` . + * Add channel lifecycle domain event + + Add support for a new domain event which can be used to track the state of + any virtio channel. It is similarl to guest agent lifecycle event. + + * **Improvements** + * vircgroupv2: Implement freezer controller + + The CGroupV2 code now supports freezer controller which allows suspending + and resuming of LXC domains. + + * virnetdevbandwidth: Only clear qdisc for defined directions + + Setting a QoS on a vNIC no longer leads to clearing (preconfigured) root + queuing discipline ('qdisc`) in both directions (egress, ingress), only the + affected one. + * **Bug fixes** * storage: ZFS: Fix incorrect volsize and refreservation on zvol creation @@ -43,6 +64,18 @@ v12.4.0 (unreleased) pre-v12.4.0 behaviour there is a new URI query parameter ``legacy_uuid`` which, if set to ``1``, will revert to the older way of working. + * virsh: Provide no auth callbacks for bash completer + + The bash completer script calls virsh to provide completions. In some cases + the underlying virsh might get stuck waiting on input (e.g. password for + SSH session). Not anymore. + + * virauth: Verify virConnectAuth::cb is set in virAuthGetPasswordPath() + + A check was missing in ``virAuthGetPasswordPath()`` which might lead to + crash of the client library when NULL callback was provided. This is now + fixed. + v12.3.0 (2026-05-02) ==================== -- 2.53.0

2 1

[PATCH v6 0/7] riscv: add initial sdext support
by Chao Liu 27 May '26

27 May '26

Hi, This v6 series is rebased on Alistair's riscv-to-apply.next branch [1]. The implementation follows RISC-V Debug Specification 1.0 [2]. It introduces the sdext/sdtrig config bits, DCSR/DPC/DSCRATCH state, Debug Mode enter/leave helpers, DRET, EBREAK entry, single-step, and trigger action=debug mode. To reduce review load, this series focuses on the Sdext features first. The Debug Module (DM) and related flows will follow in a later series. Changes in v6: - Rebased onto latest riscv-to-apply.next Changes in v5: - Patch 2: removed redundant `cpu->cfg.ext_sdext = false;` in riscv_cpu_init() since it's already set to false in MULTI_EXT_CFG_BOOL. (Daniel) - Patch 7: changed LOG_UNIMP to LOG_GUEST_ERROR for invalid trigger action=debug mode when Sdext is not enabled. LOG_UNIMP is reserved for unimplemented features, while LOG_GUEST_ERROR is for invalid guest actions. (Daniel) Changes in v4: - Fixed linux-user build errors: moved debug_mode/dcsr/dpc/dscratch fields inside #ifndef CONFIG_USER_ONLY block in cpu.h, and wrapped all code using these fields with proper guards. (Daniel) - Updated dependency from single patch (patch 5 only) to patches 1-5 of Max Chou's "Add Zvfbfa extension support" v2 series to avoid compilation errors. Changes in v3: - Rebase onto Alistair's riscv-to-apply.next branch. - Depend on Max Chou's patch "target/riscv: Use the tb->cs_bqse as the extend tb flags." (not included; apply it first) [3]. - Patch 2: default sdext to false in riscv_cpu_extensions. Changes in v2: - Drop the RHCT expected AML update from this series. - Replace the split sdext/sdtrig config bits patch with deprecating the 'debug' CPU property. - Rebase and update patch subjects to target/riscv prefix. Changes in v1: - Debug Mode entry/exit updates DCSR/DPC and restores execution via DRET. - EBREAK honors DCSR ebreak bits and enters Debug Mode when enabled. - Single-step uses DCSR.STEP with a TB flag and a helper at TB exit. It references Max Chou's patch "target/riscv: Use the tb->cs_bqse as the extend tb flags." [3]. - Sdtrig supports action=debug mode for mcontrol/mcontrol6 and reports inst-count triggers in tinfo. --- Differences vs Debug Spec (known gaps): - No Debug Module (no DMI, dmcontrol/dmstatus, haltreq/resumereq). - No debug ROM, program buffer, abstract commands, or SBA. - Resume is modeled by leaving Debug Mode at cpu_exec_enter. - Step/exception ordering is simplified: if the stepped instruction traps, the normal exception is taken and Debug Mode is not forced. - Several DCSR fields are not fully modeled (stopcount/stoptime, etc). Roadmap (next stage, DM focus): 1) Add a DM core with DMI access and hart state tracking. 2) Implement halt/resume handshake and move Debug Mode transitions under DM control. 3) Add debug ROM, program buffer, and abstract commands for GPR/CSR and memory access. 4) Add SBA if required by tooling. 5) Tighten ordering rules for step/exception/trigger priorities. CI (all jobs passed except check-patch which flags pre-existing upstream style issues, not related to this series) [4]. References: [1] https://github.com/alistair23/qemu/tree/riscv-to-apply.next [2] https://github.com/riscv/riscv-debug-spec/releases/tag/1.0 [3] https://lore.kernel.org/qemu-devel/20260108132631.9429-1-max.chou@sifive.co… [4] https://gitlab.com/TaoTang/qemu/-/pipelines/2446308835 Thanks, Chao Chao Liu (6): target/riscv: add sdext debug CSRs state target/riscv: add sdext Debug Mode helpers target/riscv: add dret instruction target/riscv: add sdext enter Debug Mode on ebreak target/riscv: add sdext single-step support target/riscv: add sdtrig trigger action=debug mode Daniel Henrique Barboza (1): target/riscv: deprecate 'debug' CPU property docs/about/deprecated.rst | 7 + include/exec/translation-block.h | 4 +- target/riscv/cpu.c | 59 +++++++- target/riscv/cpu.h | 9 ++ target/riscv/cpu_bits.h | 33 +++++ target/riscv/cpu_cfg_fields.h.inc | 3 +- target/riscv/cpu_helper.c | 90 ++++++++++++ target/riscv/csr.c | 128 +++++++++++++++++- target/riscv/debug.c | 58 +++++++- target/riscv/helper.h | 3 + target/riscv/insn32.decode | 1 + .../riscv/insn_trans/trans_privileged.c.inc | 24 ++++ target/riscv/machine.c | 44 ++++-- target/riscv/op_helper.c | 70 ++++++++++ target/riscv/tcg/tcg-cpu.c | 21 ++- target/riscv/translate.c | 15 +- 16 files changed, 544 insertions(+), 25 deletions(-) -- 2.53.0

3 11

[PATCH] tests: Link qemuxml2argvmock with test_utils_lib
by Michal Privoznik 27 May '26

27 May '26

From: Michal Privoznik <mprivozn(a)redhat.com> When running qemuxmlconftest under valgrind, it fails with a symbol lookup error: valgrind: symbol lookup error: libvirt.git/_build/tests/libqemuxml2argvmock.so: undefined symbol: virTestMakeDummyFD This occurs because qemuxml2argvmock uses the virTestMakeDummyFD() function (implemented in testutils.c) but does not explicitly link against test_utils_lib. Fix this by linking the test utils library to the mock library, statically. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- tests/meson.build | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tests/meson.build b/tests/meson.build index 83aa0104bb..d1d26fcea2 100644 --- a/tests/meson.build +++ b/tests/meson.build @@ -177,7 +177,7 @@ if conf.has('WITH_QEMU') { 'name': 'qemucapsprobemock', 'link_with': [ test_qemu_driver_lib ] }, { 'name': 'qemucpumock' }, { 'name': 'qemuhotplugmock', 'link_with': [ test_qemu_driver_lib, test_utils_qemu_lib, test_utils_lib ] }, - { 'name': 'qemuxml2argvmock' }, + { 'name': 'qemuxml2argvmock', 'link_with': [ test_utils_lib ] }, { 'name': 'virhostidmock' }, ] else -- 2.53.0

2 1

[PATCH v7 02/22] vhost: drop backend_features field
by Vladimir Sementsov-Ogievskiy 27 May '26

27 May '26

This field is mostly unused and sometimes confusing. We even have a TODO-like comment to drop it, the comment is removed in this commit. The field is used to held VHOST_USER_F_PROTOCOL_FEATURES for vhost-user and/or VHOST_NET_F_VIRTIO_NET_HDR for vhost-net (which may be vhost-user-net). But we can simply recalculate these two flags in place from hdev->features, and from net-client for VHOST_NET_F_VIRTIO_NET_HDR. Note: removing field from x-query-virtio-status result is incompatible change. We can do it because the command is unstable. Cc: devel(a)lists.libvirt.org Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov(a)yandex-team.ru> Acked-by: Markus Armbruster <armbru(a)redhat.com> Reviewed-by: Daniil Tatianin <d-tatianin(a)yandex-team.ru> Acked-by: Raphael Norwitz <raphael.s.norwitz(a)gmail.com> --- hw/block/vhost-user-blk.c | 1 - hw/net/vhost_net.c | 26 ++++++++++++++------------ hw/scsi/vhost-scsi.c | 1 - hw/scsi/vhost-user-scsi.c | 1 - hw/virtio/vdpa-dev.c | 1 - hw/virtio/vhost-user.c | 17 ++++++++--------- hw/virtio/virtio-hmp-cmds.c | 2 -- hw/virtio/virtio-qmp.c | 4 ---- include/hw/virtio/vhost.h | 7 ------- qapi/virtio.json | 3 --- 10 files changed, 22 insertions(+), 41 deletions(-) diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c index c151e836770..dc99de02635 100644 --- a/hw/block/vhost-user-blk.c +++ b/hw/block/vhost-user-blk.c @@ -348,7 +348,6 @@ static int vhost_user_blk_connect(DeviceState *dev, Error **errp) s->dev.nvqs = s->num_queues; s->dev.vqs = s->vhost_vqs; s->dev.vq_index = 0; - s->dev.backend_features = 0; vhost_dev_set_config_notifier(&s->dev, &blk_ops); diff --git a/hw/net/vhost_net.c b/hw/net/vhost_net.c index 25e9f1fd244..fda90e231e6 100644 --- a/hw/net/vhost_net.c +++ b/hw/net/vhost_net.c @@ -52,8 +52,14 @@ int vhost_net_set_config(struct vhost_net *net, const uint8_t *data, void vhost_net_ack_features_ex(struct vhost_net *net, const uint64_t *features) { - virtio_features_copy(net->dev.acked_features_ex, - net->dev.backend_features_ex); + virtio_features_clear(net->dev.acked_features_ex); + if (net->backend == -1) { + net->dev.acked_features = + net->dev.features & (1ULL << VHOST_USER_F_PROTOCOL_FEATURES); + } else if (!qemu_has_vnet_hdr(net->nc)) { + net->dev.acked_features = 1ULL << VHOST_NET_F_VIRTIO_NET_HDR; + } + vhost_ack_features_ex(&net->dev, net->feature_bits, features); } @@ -258,12 +264,9 @@ struct vhost_net *vhost_net_init(VhostNetOptions *options) if (r < 0) { goto fail; } - net->dev.backend_features = qemu_has_vnet_hdr(options->net_backend) - ? 0 : (1ULL << VHOST_NET_F_VIRTIO_NET_HDR); net->backend = r; net->dev.protocol_features = 0; } else { - virtio_features_clear(net->dev.backend_features_ex); net->dev.protocol_features = 0; net->backend = -1; @@ -284,13 +287,12 @@ struct vhost_net *vhost_net_init(VhostNetOptions *options) net->dev.features &= ~(1ULL << VIRTIO_NET_F_MRG_RXBUF); } - if (virtio_features_andnot(missing_features, - net->dev.backend_features_ex, - net->dev.features_ex)) { - fprintf(stderr, "vhost lacks feature mask 0x" VIRTIO_FEATURES_FMT - " for backend\n", VIRTIO_FEATURES_PR(missing_features)); - goto fail; - } + if (!qemu_has_vnet_hdr(options->net_backend) && + (~net->dev.features & (1ULL << VHOST_NET_F_VIRTIO_NET_HDR))) { + fprintf(stderr, "vhost lacks feature mask 0x%llx for backend\n", + ~net->dev.features & (1ULL << VHOST_NET_F_VIRTIO_NET_HDR)); + goto fail; + } } /* Set sane init value. Override when guest acks. */ diff --git a/hw/scsi/vhost-scsi.c b/hw/scsi/vhost-scsi.c index 699863cc109..859888fdbaf 100644 --- a/hw/scsi/vhost-scsi.c +++ b/hw/scsi/vhost-scsi.c @@ -275,7 +275,6 @@ static void vhost_scsi_realize(DeviceState *dev, Error **errp) vqs = g_new0(struct vhost_virtqueue, vsc->dev.nvqs); vsc->dev.vqs = vqs; vsc->dev.vq_index = 0; - vsc->dev.backend_features = 0; ret = vhost_dev_init(&vsc->dev, (void *)(uintptr_t)vhostfd, VHOST_BACKEND_TYPE_KERNEL, 0, errp); diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c index 3612897d4b9..5070178dc28 100644 --- a/hw/scsi/vhost-user-scsi.c +++ b/hw/scsi/vhost-user-scsi.c @@ -159,7 +159,6 @@ static int vhost_user_scsi_connect(DeviceState *dev, Error **errp) vsc->dev.nvqs = VIRTIO_SCSI_VQ_NUM_FIXED + vs->conf.num_queues; vsc->dev.vqs = s->vhost_vqs; vsc->dev.vq_index = 0; - vsc->dev.backend_features = 0; ret = vhost_dev_init(&vsc->dev, &s->vhost_user, VHOST_BACKEND_TYPE_USER, 0, errp); diff --git a/hw/virtio/vdpa-dev.c b/hw/virtio/vdpa-dev.c index 4532d63653e..06179e402e4 100644 --- a/hw/virtio/vdpa-dev.c +++ b/hw/virtio/vdpa-dev.c @@ -104,7 +104,6 @@ static void vhost_vdpa_device_realize(DeviceState *dev, Error **errp) v->dev.vqs = vqs; v->dev.vq_index = 0; v->dev.vq_index_end = v->dev.nvqs; - v->dev.backend_features = 0; v->started = false; ret = vhost_vdpa_get_iova_range(v->vhostfd, &iova_range); diff --git a/hw/virtio/vhost-user.c b/hw/virtio/vhost-user.c index 13c73e7b178..7ce17c5b2d8 100644 --- a/hw/virtio/vhost-user.c +++ b/hw/virtio/vhost-user.c @@ -1465,14 +1465,15 @@ static int vhost_user_set_features(struct vhost_dev *dev, int ret; /* - * We need to include any extra backend only feature bits that - * might be needed by our device. Currently this includes the - * VHOST_USER_F_PROTOCOL_FEATURES bit for enabling protocol - * features. + * Don't lose VHOST_USER_F_PROTOCOL_FEATURES, which is vhost-user + * specific. */ - ret = vhost_user_set_u64(dev, VHOST_USER_SET_FEATURES, - features | dev->backend_features, - log_enabled); + if (virtio_has_feature(dev->features, VHOST_USER_F_PROTOCOL_FEATURES)) { + features |= 1ULL << VHOST_USER_F_PROTOCOL_FEATURES; + } + + ret = vhost_user_set_u64(dev, VHOST_USER_SET_FEATURES, features, + log_enabled); if (virtio_has_feature(dev->protocol_features, VHOST_USER_PROTOCOL_F_STATUS)) { @@ -2192,8 +2193,6 @@ static int vhost_user_backend_init(struct vhost_dev *dev, void *opaque, (dev->config_ops && dev->config_ops->vhost_dev_config_notifier); uint64_t protocol_features; - dev->backend_features |= 1ULL << VHOST_USER_F_PROTOCOL_FEATURES; - err = vhost_user_get_u64(dev, VHOST_USER_GET_PROTOCOL_FEATURES, &protocol_features); if (err < 0) { diff --git a/hw/virtio/virtio-hmp-cmds.c b/hw/virtio/virtio-hmp-cmds.c index 1daae482d37..4bf9a3109d4 100644 --- a/hw/virtio/virtio-hmp-cmds.c +++ b/hw/virtio/virtio-hmp-cmds.c @@ -176,8 +176,6 @@ void hmp_virtio_status(Monitor *mon, const QDict *qdict) hmp_virtio_dump_features(mon, s->vhost_dev->features); monitor_printf(mon, " Acked features:\n"); hmp_virtio_dump_features(mon, s->vhost_dev->acked_features); - monitor_printf(mon, " Backend features:\n"); - hmp_virtio_dump_features(mon, s->vhost_dev->backend_features); monitor_printf(mon, " Protocol features:\n"); hmp_virtio_dump_protocols(mon, s->vhost_dev->protocol_features); } diff --git a/hw/virtio/virtio-qmp.c b/hw/virtio/virtio-qmp.c index f9cdca50d99..71717c56a41 100644 --- a/hw/virtio/virtio-qmp.c +++ b/hw/virtio/virtio-qmp.c @@ -710,8 +710,6 @@ VirtioStatus *qmp_x_query_virtio_status(const char *path, Error **errp) vdev->guest_features_ex); status->host_features = qmp_decode_features(vdev->device_id, vdev->host_features_ex); - status->backend_features = qmp_decode_features(vdev->device_id, - vdev->backend_features_ex); switch (vdev->device_endian) { case VIRTIO_DEVICE_ENDIAN_LITTLE: @@ -752,8 +750,6 @@ VirtioStatus *qmp_x_query_virtio_status(const char *path, Error **errp) qmp_decode_features(vdev->device_id, hdev->features_ex); status->vhost_dev->acked_features = qmp_decode_features(vdev->device_id, hdev->acked_features_ex); - status->vhost_dev->backend_features = - qmp_decode_features(vdev->device_id, hdev->backend_features_ex); status->vhost_dev->protocol_features = qmp_decode_protocols(hdev->protocol_features); diff --git a/include/hw/virtio/vhost.h b/include/hw/virtio/vhost.h index 3dfec5e15cc..e139e059857 100644 --- a/include/hw/virtio/vhost.h +++ b/include/hw/virtio/vhost.h @@ -100,16 +100,9 @@ struct vhost_dev { * * @features: available features provided by the backend * @acked_features: final negotiated features with front-end driver - * - * @backend_features: this is used in a couple of places to either - * store VHOST_USER_F_PROTOCOL_FEATURES to apply to - * VHOST_USER_SET_FEATURES or VHOST_NET_F_VIRTIO_NET_HDR. Its - * future use should be discouraged and the variable retired as - * its easy to confuse with the VirtIO backend_features. */ VIRTIO_DECLARE_FEATURES(features); VIRTIO_DECLARE_FEATURES(acked_features); - VIRTIO_DECLARE_FEATURES(backend_features); /** * @protocol_features: is the vhost-user only feature set by diff --git a/qapi/virtio.json b/qapi/virtio.json index 09dd0e6d05a..671f1ad793c 100644 --- a/qapi/virtio.json +++ b/qapi/virtio.json @@ -85,8 +85,6 @@ # # @acked-features: vhost_dev acked_features # -# @backend-features: vhost_dev backend_features -# # @protocol-features: vhost_dev protocol_features # # @max-queues: vhost_dev max_queues @@ -106,7 +104,6 @@ 'vq-index': 'int', 'features': 'VirtioDeviceFeatures', 'acked-features': 'VirtioDeviceFeatures', - 'backend-features': 'VirtioDeviceFeatures', 'protocol-features': 'VhostDeviceProtocols', 'max-queues': 'uint64', 'backend-cap': 'uint64', -- 2.52.0

1 0

[PATCH v1 0/7] Add TLS-PSK support for migration
by Abhisek Panda 26 May '26

26 May '26

QEMU provides the ability to encrypt the migration data stream using two transport layer security (TLS) authentication schemes: X.509 certificates and pre-shared keys (PSK). Currently, Libvirt only supports the X.509-based TLS authentication scheme. In TLS X.509 certificates, a set of live migrations utilize a fixed set of static certificates for encrypted migration. In this authentication scheme, users require to deploy a certificate authority and monitor the certificate expiration window. In case certificates are compromised all the future live migrations are vulnerable. To resolve this, the patch series introduce the alternative pre-shared key-based authentication scheme. This mechanism relies on symmetric pre-shared keys (a secret key that is known to both sender and receiver prior to secure communication) for providing secure transfer of data. During a migration, the parties negotiate which unique key to utilize for encrypting the migration data. To improve the security further, we utilize different key files for each migration session. Abhisek Panda (7): conf: Add configuration params for TLS-PSK include: define VIR_MIGRATE_TLS_PSK flag qemu: Add support to build the tls-creds-psk object qemu: rename tls-creds-x509 obj related functions qemu: Manage tls-creds-psk object lifecycle qemu: Set up the migrate TLS-PSK objects include: define VIR_MIGRATE_PARAM_TLS_PSK_DIRECTORY include/libvirt/libvirt-domain.h | 31 ++- src/qemu/libvirtd_qemu.aug | 2 + src/qemu/qemu.conf.in | 27 ++- src/qemu/qemu_alias.c | 19 +- src/qemu/qemu_alias.h | 5 +- src/qemu/qemu_backup.c | 2 +- src/qemu/qemu_command.c | 31 ++- src/qemu/qemu_command.h | 8 + src/qemu/qemu_conf.c | 55 ++++- src/qemu/qemu_conf.h | 3 + src/qemu/qemu_domain.c | 2 +- src/qemu/qemu_driver.c | 24 ++- src/qemu/qemu_hotplug.c | 125 +++++++---- src/qemu/qemu_hotplug.h | 43 ++-- src/qemu/qemu_migration.c | 194 ++++++++++++------ src/qemu/qemu_migration.h | 3 + src/qemu/qemu_migration_params.c | 138 ++++++++++--- src/qemu/qemu_migration_params.h | 28 ++- src/qemu/qemu_postparse.c | 2 +- src/qemu/test_libvirtd_qemu.aug.in | 2 + tests/qemumigparamsdata/tls-enabled.json | 2 +- tests/qemumigparamsdata/tls-enabled.reply | 2 +- tests/qemumigparamsdata/tls-enabled.xml | 2 +- tests/qemumigparamsdata/tls-hostname.json | 2 +- tests/qemumigparamsdata/tls-hostname.reply | 2 +- tests/qemumigparamsdata/tls-hostname.xml | 2 +- tests/qemumonitorjsontest.c | 4 +- tests/qemustatusxml2xmldata/upgrade-out.xml | 2 +- .../chardev-backends-json.x86_64-9.1.0.args | 8 +- .../chardev-backends-json.x86_64-latest.args | 8 +- .../chardev-backends.x86_64-9.1.0.args | 8 +- .../chardev-backends.x86_64-latest.args | 8 +- ...rk-tlsx509-nbd-hostname.x86_64-latest.args | 6 +- ...isk-network-tlsx509-nbd.x86_64-latest.args | 6 +- ...-tlsx509-chardev-verify.x86_64-latest.args | 4 +- ...ial-tcp-tlsx509-chardev.x86_64-latest.args | 4 +- ...-tlsx509-secret-chardev.x86_64-latest.args | 4 +- tests/testutilsqemu.c | 2 + tools/virsh-domain.c | 12 ++ 39 files changed, 623 insertions(+), 209 deletions(-) -- 2.39.3

1 7