January 2026 - Devel - libvirt List Archives

[PATCH] qemu: Add stop hook support for domain shutdown
by Mitchel Humpherys 06 Jan '26

06 Jan '26

Introduce a new QEMU hook operation "stop" that is called before a domain is terminated (via virsh shutdown or virsh destroy). This allows external scripts to perform cleanup or veto the stop process by returning a non-zero exit code. The hook is called as: /etc/libvirt/hooks/qemu <guest_name> stop begin - The full domain XML is provided on stdin. Example qemu hook that protects domains from being stopped (to be dropped in /etc/libfirt/hooks/qemu.d/10-vm-protection or similar): ```bash #!/bin/bash # Protection hook - vetos domain stops for domains listed in # /etc/libvirt/protected-domains set -euo pipefail readonly PROTECT_DIR="/etc/libvirt" readonly DOMAIN_NAME="${1:-}" readonly OPERATION="${2:-}" # Only act on stop operation if [[ "$OPERATION" != "stop" ]]; then exit 0 fi readonly PROTECT_LIST="${PROTECT_DIR}/protected-domains" if [[ -f "$PROTECT_LIST" ]]; then # Read file, strip comments and empty lines, check for exact domain match if grep -Fxq -- "$DOMAIN_NAME" <(sed -e 's/#.*$//' -e '/^[[:space:]]*$/d' "$PROTECT_LIST"); then logger -t libvirt-hook-test "20-protection: BLOCKING stop for protected domain '$DOMAIN_NAME'" echo "$(date '+%Y-%m-%d %H:%M:%S'): BLOCKED stop for '$DOMAIN_NAME'" \ >> /tmp/libvirt-test-logs/protection.log printf 'vm-protection: stop blocked for %s\n' "$DOMAIN_NAME" >&2 exit 1 fi fi exit 0 ``` With a list of domains to protect in /etc/libvirt/protected-domains. With the above hook in place, attempting to destroy a domain listed in /etc/libvirt/protected-domains will fail: ``` > virsh list --all Id Name State ------------------------------- 2 protected-vm running - test-vm shut off > virsh destroy protected-vm error: Failed to destroy domain 'protected-vm' error: Hook script execution failed: internal error: Child process (LC_ALL=C LD_LIBRARY_PATH=/opt/libvirt-test/lib/x86_64-linux-gnu:/opt/libvirt-test/lib PATH=/usr/local/sbin:/usr/local/bin:/usr/bin HOME=/root USER=root LOGNAME=root /opt/libvirt-test/etc/libvirt/hooks/qemu.d/20-protection protected-vm stop begin -) unexpected exit status 1: vm-protection: stop blocked for protected-vm > virsh list --all Id Name State ------------------------------- 2 protected-vm running - test-vm shut off ``` Signed-off-by: Mitchel Humpherys <mitch.special(a)gmail.com> --- docs/hooks.rst | 14 +++++++++++--- src/qemu/qemu_process.c | 17 +++++++++++++++++ src/util/virhook.c | 1 + src/util/virhook.h | 1 + 4 files changed, 30 insertions(+), 3 deletions(-) diff --git a/docs/hooks.rst b/docs/hooks.rst index e1745b8cc7..5ae03a36cd 100644 --- a/docs/hooks.rst +++ b/docs/hooks.rst @@ -202,9 +202,17 @@ operation. There is no specific operation to indicate a "restart" is occurring. /etc/libvirt/hooks/qemu guest_name started begin - -- When a QEMU guest is stopped, the qemu hook script is called in two - locations, to match the startup. First, :since:`since 0.8.0`, the hook is - called before libvirt restores any labels: +- When a QEMU guest is stopped, the qemu hook script is called in three + locations, to match the startup. The first location, :since:`since 11.10.0`, + is called before the domain is stopped. This allows the hook to perform + cleanup tasks or veto the stop operation by returning a non-zero exit code: + + :: + + /etc/libvirt/hooks/qemu guest_name stop begin - + + The second location, :since:`since 0.8.0`, is called after the QEMU process + has terminated but before libvirt restores any labels: :: diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index 0e50cd1ccc..fa2b691d99 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -9015,6 +9015,23 @@ qemuProcessBeginStopJob(virDomainObj *vm, qemuDomainObjPrivate *priv = vm->privateData; unsigned int killFlags = forceKill ? VIR_QEMU_PROCESS_KILL_FORCE : 0; + /* call stop hook if present */ + if (virHookPresent(VIR_HOOK_DRIVER_QEMU)) { + virQEMUDriver *driver = priv->driver; + g_autofree char *xml = qemuDomainDefFormatXML(driver, NULL, vm->def, 0); + int hookret; + + if (!xml) + return -1; + + hookret = virHookCall(VIR_HOOK_DRIVER_QEMU, vm->def->name, + VIR_HOOK_QEMU_OP_STOP, VIR_HOOK_SUBOP_BEGIN, + NULL, xml, NULL); + + if (hookret < 0) + return -1; + } + /* We need to prevent monitor EOF callback from doing our work (and * sending misleading events) while the vm is unlocked inside * BeginJob/ProcessKill API or any other code path before 'vm->def->id' is diff --git a/src/util/virhook.c b/src/util/virhook.c index d012bb1825..01ba17e406 100644 --- a/src/util/virhook.c +++ b/src/util/virhook.c @@ -76,6 +76,7 @@ VIR_ENUM_IMPL(virHookSubop, VIR_ENUM_IMPL(virHookQemuOp, VIR_HOOK_QEMU_OP_LAST, "start", + "stop", "stopped", "prepare", "release", diff --git a/src/util/virhook.h b/src/util/virhook.h index d8237c837e..ea8c540c3f 100644 --- a/src/util/virhook.h +++ b/src/util/virhook.h @@ -52,6 +52,7 @@ typedef enum { typedef enum { VIR_HOOK_QEMU_OP_START, /* domain is about to start */ + VIR_HOOK_QEMU_OP_STOP, /* domain is about to stop */ VIR_HOOK_QEMU_OP_STOPPED, /* domain has stopped */ VIR_HOOK_QEMU_OP_PREPARE, /* domain startup initiated */ VIR_HOOK_QEMU_OP_RELEASE, /* domain destruction is over */ -- 2.52.0

2 2

[PATCH v2 0/2] qemu: Relax hyperv features validation
by Michal Privoznik 06 Jan '26

06 Jan '26

v2 of: https://lists.libvirt.org/archives/list/devel@lists.libvirt.org/thread/5FLY… diff to v1: - Patch 1/3 from original series is dropped, as I misunderstood QEMU's code and though Q35 is not affected. It is. - Patch 2/3 from original series is reworked to drop the check entirely, again, I blame my code misunderstanding. Michal Prívozník (2): qemu_validate: Drop VIR_DOMAIN_HYPERV_SYNIC dependency on VIR_DOMAIN_HYPERV_VPINDEX qemu_validate: Drop VIR_DOMAIN_HYPERV_STIMER dependency on VIR_DOMAIN_HYPERV_VPINDEX src/qemu/qemu_validate.c | 3 --- 1 file changed, 3 deletions(-) -- 2.52.0

2 3

[PATCH 0/3] qemu: Relax hyperv features validation
by Michal Privoznik 06 Jan '26

06 Jan '26

*** BLURB HERE *** Michal Prívozník (3): qemuxmlconfdata: Switch some hyperv tests from i440fx to q35 qemu: Skip validation of VIR_DOMAIN_HYPERV_SYNIC on i440fx qemu_validate: Drop VIR_DOMAIN_HYPERV_STIMER dependency on VIR_DOMAIN_HYPERV_VPINDEX src/qemu/qemu_validate.c | 9 +++++-- .../hyperv-host-model.x86_64-latest.args | 9 +++++-- .../hyperv-host-model.x86_64-latest.xml | 24 ++++++++++++++++--- tests/qemuxmlconfdata/hyperv-host-model.xml | 4 ++-- .../qemuxmlconfdata/hyperv.x86_64-latest.args | 9 +++++-- .../qemuxmlconfdata/hyperv.x86_64-latest.xml | 24 ++++++++++++++++--- tests/qemuxmlconfdata/hyperv.xml | 4 ++-- 7 files changed, 67 insertions(+), 16 deletions(-) -- 2.52.0

3 7

[PATCH 00/10] qemu: Snapshot and saveimage related refactors (background snapshot saga part 2)
by Peter Krempa 06 Jan '26

06 Jan '26

This series fixes up some code (mostly removes unneeded) 'driver' argument and fixes control flow in some snapshot and save image related functions. This series also contains 2 patches necromanced from the original attempt to add background snapshot from Nikolay. Nikolay Shirokovskiy (2): qemu: snapshot: Don't overwrite 'ret' in qemuSnapshotCreateActiveExternal qemu: snapshot: Check earlier if memory snapshot is possible Peter Krempa (8): qemuHotplugAttachDBusVMState: Remove 'driver' argument qemuMigrationSrcToLegacyFile: Remove 'driver' argument qemuDomainEventEmitJobCompleted: Remove 'driver' argument qemuMigrationSrcToSparseFile: Remove 'driver' argument qemuMigrationSrcToFile: Remove 'driver' argument qemuSaveImageCreateFd: Remove 'driver' argument qemuSaveImageCreate: Remove 'driver' argument virQEMUSaveDataNew: Refactor cleanup src/qemu/qemu_backup.c | 2 +- src/qemu/qemu_domainjob.c | 6 +++--- src/qemu/qemu_domainjob.h | 3 +-- src/qemu/qemu_driver.c | 4 ++-- src/qemu/qemu_hotplug.c | 6 ++---- src/qemu/qemu_hotplug.h | 3 +-- src/qemu/qemu_migration.c | 36 ++++++++++++++++++------------------ src/qemu/qemu_migration.h | 5 ++--- src/qemu/qemu_saveimage.c | 27 ++++++++++++--------------- src/qemu/qemu_saveimage.h | 3 +-- src/qemu/qemu_snapshot.c | 22 ++++++++++++---------- 11 files changed, 55 insertions(+), 62 deletions(-) -- 2.52.0

3 13

[PATCH 0/2] qemu: capabilities: Update '10.2' capabilities on x86_64 after release
by Peter Krempa 06 Jan '26

06 Jan '26

Peter Krempa (2): qemucapabilitiesdata: Re-generate 'caps_10.2.0_x86_64' qemu: capabilities: Update '10.2' capabilities on x86_64 after release .../domaincapsdata/qemu_10.2.0-q35.x86_64.xml | 170 +++++++++++-- tests/domaincapsdata/qemu_10.2.0.x86_64.xml | 170 +++++++++++-- .../caps_10.2.0_x86_64.replies | 230 +++++++++++++----- .../caps_10.2.0_x86_64.xml | 157 ++++++++++-- ...host-model-fallback-kvm.x86_64-latest.args | 2 +- ...cpu-host-model-features.x86_64-latest.args | 2 +- .../cpu-host-model-kvm.x86_64-latest.args | 2 +- ...st-model-nofallback-kvm.x86_64-latest.args | 2 +- 8 files changed, 618 insertions(+), 117 deletions(-) -- 2.52.0

2 3

[PATCH v4 0/9] Enable mshv support for Qemu domains
by Praveen K Paladugu 06 Jan '26

06 Jan '26

This patch set introduces changes within libvirt to lauch Qemu guests with Microsoft Hypervisor(mshv). Qemu exposes this capability as 'mshv' accelerator type. Below is an excerpt of domain XML with mshv accelerator.: ``` <domain type='hyperv'> <name>qemu_mshv_test</name> <os> <type machine='q35'>hvm</type> </os> ... ``` 'mshv' accelerator within Qemu translates to 'hyperv' domains in Libvirt. Changs in v4: - Adopted `query-accelerators` API to get mshv capabilities. - Fixed virtytpe parsing in schemas. Chages in v3: - Addressed review comments on v2. - fixed qemucapabilitiestest case - Added a case to qemuxmlconftest Changes in v2: - Addressed review comments on v1. - Added a test to verify mshv capabilities. Eugene Fedorenko (1): qemu: Append mshv to hwaccel list. Praveen K Paladugu (8): qemu: Introduce mshv capabilities qemu: Query and use mshv capabilities qemu: use mshv accel for hyperv domains qemu: conditionally add /dev/mshv to acl conf: Fix virttype handling in schemas tests: Add domaincaps case for mshv tests: qemucapabilities: Introduce MSHV capability tests:qemuxmlconftest: Add case for hyperv domains src/conf/schemas/basictypes.rng | 22 + src/conf/schemas/capability.rng | 14 +- src/conf/schemas/domaincommon.rng | 23 +- src/qemu/qemu_capabilities.c | 39 +- src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_cgroup.c | 6 + src/qemu/qemu_command.c | 4 +- src/qemu/qemu_namespace.c | 2 + src/qemu/qemu_process.c | 2 +- .../qemu_10.2.0-q35.x86_64+mshv.xml | 222 + .../qemu_10.2.0-tcg.x86_64+mshv.xml | 2334 + .../qemu_10.2.0.x86_64+mshv.xml | 222 + .../caps_10.2.0_x86_64+mshv.replies | 45176 ++++++++++++++++ .../caps_10.2.0_x86_64+mshv.xml | 3212 ++ .../qemucaps2xmloutdata/caps.x86_64+mshv.xml | 29 + ...86_64-q35-headless.x86_64-latest+mshv.args | 48 + ...x86_64-q35-headless.x86_64-latest+mshv.xml | 87 + ...mshv-x86_64-q35-headless.x86_64-latest.err | 1 + .../mshv-x86_64-q35-headless.xml | 35 + tests/qemuxmlconftest.c | 5 + 20 files changed, 51446 insertions(+), 38 deletions(-) create mode 100644 tests/domaincapsdata/qemu_10.2.0-q35.x86_64+mshv.xml create mode 100644 tests/domaincapsdata/qemu_10.2.0-tcg.x86_64+mshv.xml create mode 100644 tests/domaincapsdata/qemu_10.2.0.x86_64+mshv.xml create mode 100644 tests/qemucapabilitiesdata/caps_10.2.0_x86_64+mshv.replies create mode 100644 tests/qemucapabilitiesdata/caps_10.2.0_x86_64+mshv.xml create mode 100644 tests/qemucaps2xmloutdata/caps.x86_64+mshv.xml create mode 100644 tests/qemuxmlconfdata/mshv-x86_64-q35-headless.x86_64-latest+mshv.args create mode 100644 tests/qemuxmlconfdata/mshv-x86_64-q35-headless.x86_64-latest+mshv.xml create mode 100644 tests/qemuxmlconfdata/mshv-x86_64-q35-headless.x86_64-latest.err create mode 100644 tests/qemuxmlconfdata/mshv-x86_64-q35-headless.xml -- 2.51.0

4 24

[PATCH] libvirt-guests.service: use Type=exec
by Antonio Terceiro 06 Jan '26

06 Jan '26

Type=oneshot means that the service is only considered started after the main process exits. In setups where libvirt-guests is configured to be used and imposes a delay on the startup of guests, this will cause systemd to report a quite high boot time as it will wait until all guests have booted, plus all the delay between them. High boot times are often reported as issues in monitoring systems, and require operator attention, while in this case there is not really an issue with the system, it's just doing what it was configured to do. With Type=exec, the service is considered started right after the main process is *started*, but the boot process will not wait for it. Having other units depend on libvirt-guests.service finishing was already not reliable because one cannot know how long the OS inside each guest takes to boot, and libvirt-guests already does not wait for them to be in any particular state. Signed-off-by: Antonio Terceiro <terceiro(a)debian.org> --- tools/libvirt-guests.service.in | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/libvirt-guests.service.in b/tools/libvirt-guests.service.in index b044444298..c8e4d3d749 100644 --- a/tools/libvirt-guests.service.in +++ b/tools/libvirt-guests.service.in @@ -19,7 +19,7 @@ EnvironmentFile=-@initconfdir@/libvirt-guests # out the code ExecStart=@libexecdir@/libvirt-guests.sh start ExecStop=@libexecdir@/libvirt-guests.sh stop -Type=oneshot +Type=exec RemainAfterExit=yes StandardOutput=journal+console TimeoutStopSec=0 -- 2.51.0

3 4

[PATCH] docs: fix broken UUID example
by Daniel P. Berrangé 06 Jan '26

06 Jan '26

From: Daniel P. Berrangé <berrange(a)redhat.com> The UUID had only 11 characters in the last component, instead of the required 12. Signed-off-by: Daniel P. Berrangé <berrange(a)redhat.com> --- docs/formatsecret.rst | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/docs/formatsecret.rst b/docs/formatsecret.rst index d4e0cb9ddb..bbf2fc22ea 100644 --- a/docs/formatsecret.rst +++ b/docs/formatsecret.rst @@ -73,7 +73,7 @@ passphrase to decrypt the volume, :since:`since 2.1.0`. An example follows: # cat luks-secret.xml <secret ephemeral='no' private='yes'> <description>LUKS Sample Secret</description> - <uuid>f52a81b2-424e-490c-823d-6bd4235bc57</uuid> + <uuid>f52a81b2-424e-490c-823d-6bd4235bc507</uuid> <usage type='volume'> <volume>/var/lib/libvirt/images/luks-sample.img</volume> </usage> @@ -91,7 +91,7 @@ The volume type secret can be supplied in domain XML for a luks storage volume :: <encryption format='luks'> - <secret type='passphrase' uuid='f52a81b2-424e-490c-823d-6bd4235bc57'/> + <secret type='passphrase' uuid='f52a81b2-424e-490c-823d-6bd4235bc507'/> </encryption> Usage type "ceph" -- 2.52.0

2 2

[PATCH 00/36] qemu: Improvements and fixes to firmware selection
by Andrea Bolognani 06 Jan '26

06 Jan '26

This series improves validation so that more nonsensical configurations are rejected, fixes a number of scenarios in which user-provided attributes were getting overwritten by the firmware selection process, and overall makes things more predictable and reliable. Notably, it addresses the inability of starting confidential VMs on aarch64, which was reported[1] some time ago. It is also a prerequisite of another series that I will post shortly, which introduces support for the uefi-vars QEMU device and thus makes it possible to use Secure Boot for aarch64 VMs. Since all these fixes and improvements make sense on their own, and there is a little bit of work still needed on the QEMU/edk2 side before the other series can be merged, I decided to post this one separately instead of lumping them together. It's not like it's not meaty enough on its own anyway :) [1] https://lists.libvirt.org/archives/list/devel@lists.libvirt.org/thread/D5UD… Andrea Bolognani (36): qemu_firmware: Drop support for kernel descriptors qemu_firmware: Drop 'nvram' local variable qemu_firmware: Move format=raw compat exception qemu_firmware: Move copying of nvram.format to loader.format tests: Add firmware-manual-efi-rw-nvram domain_validate: Reject NVRAM with read/write firmware tests: Add firmware-auto-bios-rw tests: Add firmware-manual-bios-rw domain_validate: Reject read/write ROMs tests: Add firmware-auto-efi-format-loader-qcow2-rom domain_validate: Reject ROMs with format other than raw qemu_firmware: Ignore stateless/combined when NVRAM is configured qemu_firmware: Drop fallback for absent nvramTemplateFormat schemas: Allow templateFormat without template path tests: Add firmware-manual-efi-nvram-template-nonstandard-format tests: Add firmware-manual-efi-nvram-template-nonstandard-legacy-paths tests: Add firmware-auto-efi-format-nvram-raw tests: Add firmware-auto-efi-format-nvram-raw-loader-path tests: Add firmware-auto-efi-format-nvram-raw-nvramtemplate-path tests: Add firmware-auto-efi-format-nvramtemplate-qcow2 tests: Add firmware-auto-efi-format-mismatch-nvramtemplate qemu_firmware: Introduce qemuFirmwareFillDomainCustom() qemu_firmware: Set templateFormat for custom paths qemu_firmware: Simplify handling of legacy paths qemu_firmware: Refactor setting NVRAM format qemu_firmware: Prefer template format to loader format qemu_firmware: Retain user-specified NVRAM format qemu_firmware: Take templateFormat into account when matching qemu_firmware: Take NVRAM format into account when matching qemu_firmware: Remove NVRAM to loader format copy hack tests: Add firmware-manual-efi-sev-snp tests: Add firmware-manual-efi-tdx qemu_firmware: ROM firmware is always in raw format qemu_firmware: Don't skip autoselection for ROM qemu_firmware: Allow matching both UEFI and BIOS for ROM loader news: Mention improvements and fixes to firmware selection NEWS.rst | 8 + src/conf/domain_conf.c | 18 +- src/conf/domain_validate.c | 30 ++ src/conf/schemas/domaincommon.rng | 10 +- src/qemu/qemu_firmware.c | 367 ++++++++++-------- src/qemu/qemu_postparse.c | 17 - .../firmware-auto-bios-rw.x86_64-latest.err | 1 + ...> firmware-auto-bios-rw.x86_64-latest.xml} | 5 +- .../qemuxmlconfdata/firmware-auto-bios-rw.xml | 18 + ...-format-loader-qcow2-rom.x86_64-latest.err | 1 + ...mware-auto-efi-format-loader-qcow2-rom.xml | 18 + ...t-mismatch-nvramtemplate.x86_64-latest.err | 1 + ...-mismatch-nvramtemplate.x86_64-latest.xml} | 6 +- ...auto-efi-format-mismatch-nvramtemplate.xml | 19 + ...uto-efi-format-mismatch.x86_64-latest.args | 5 +- ...auto-efi-format-mismatch.x86_64-latest.xml | 2 +- ...-nvram-raw-loader-path.x86_64-latest.args} | 4 +- ...t-nvram-raw-loader-path.x86_64-latest.xml} | 4 +- ...-auto-efi-format-nvram-raw-loader-path.xml | 19 + ...raw-nvramtemplate-path.x86_64-latest.args} | 4 +- ...-raw-nvramtemplate-path.x86_64-latest.xml} | 4 +- ...fi-format-nvram-raw-nvramtemplate-path.xml | 18 + ...t-nvram-raw.x86_64-latest.abi-update.args} | 0 ...at-nvram-raw.x86_64-latest.abi-update.xml} | 0 ...o-efi-format-nvram-raw.x86_64-latest.args} | 0 ...to-efi-format-nvram-raw.x86_64-latest.xml} | 0 .../firmware-auto-efi-format-nvram-raw.xml | 18 + ...at-nvramtemplate-qcow2.x86_64-latest.args} | 9 +- ...mat-nvramtemplate-qcow2.x86_64-latest.xml} | 4 +- ...re-auto-efi-format-nvramtemplate-qcow2.xml | 18 + .../firmware-manual-bios-rw.x86_64-latest.err | 1 + .../firmware-manual-bios-rw.xml | 15 + ...-loader-path-nonstandard.x86_64-latest.xml | 2 +- ...ate-nonstandard-format.x86_64-latest.args} | 10 +- ...late-nonstandard-format.x86_64-latest.xml} | 4 +- ...-efi-nvram-template-nonstandard-format.xml | 19 + ...nstandard-legacy-paths.x86_64-latest.args} | 4 +- ...onstandard-legacy-paths.x86_64-latest.xml} | 5 +- ...vram-template-nonstandard-legacy-paths.xml | 20 + ...ram-template-nonstandard.x86_64-latest.xml | 2 +- ...ware-manual-efi-rw-nvram.x86_64-latest.err | 1 + .../firmware-manual-efi-rw-nvram.xml | 19 + ...ual-efi-sev-snp.x86_64-latest+amdsev.args} | 7 +- ...nual-efi-sev-snp.x86_64-latest+amdsev.xml} | 12 +- .../firmware-manual-efi-sev-snp.xml | 21 + ...anual-efi-tdx.x86_64-latest+inteltdx.args} | 9 +- ...manual-efi-tdx.x86_64-latest+inteltdx.xml} | 11 +- .../firmware-manual-efi-tdx.xml | 25 ++ tests/qemuxmlconftest.c | 19 + 49 files changed, 571 insertions(+), 263 deletions(-) create mode 100644 tests/qemuxmlconfdata/firmware-auto-bios-rw.x86_64-latest.err copy tests/qemuxmlconfdata/{firmware-manual-efi-nvram-template-nonstandard.x86_64-latest.xml => firmware-auto-bios-rw.x86_64-latest.xml} (83%) create mode 100644 tests/qemuxmlconfdata/firmware-auto-bios-rw.xml create mode 100644 tests/qemuxmlconfdata/firmware-auto-efi-format-loader-qcow2-rom.x86_64-latest.err create mode 100644 tests/qemuxmlconfdata/firmware-auto-efi-format-loader-qcow2-rom.xml create mode 100644 tests/qemuxmlconfdata/firmware-auto-efi-format-mismatch-nvramtemplate.x86_64-latest.err copy tests/qemuxmlconfdata/{firmware-manual-efi-nvram-template-nonstandard.x86_64-latest.xml => firmware-auto-efi-format-mismatch-nvramtemplate.x86_64-latest.xml} (83%) create mode 100644 tests/qemuxmlconfdata/firmware-auto-efi-format-mismatch-nvramtemplate.xml copy tests/qemuxmlconfdata/{firmware-auto-efi-format-mismatch.x86_64-latest.args => firmware-auto-efi-format-nvram-raw-loader-path.x86_64-latest.args} (90%) copy tests/qemuxmlconfdata/{firmware-auto-efi-format-mismatch.x86_64-latest.xml => firmware-auto-efi-format-nvram-raw-loader-path.x86_64-latest.xml} (81%) create mode 100644 tests/qemuxmlconfdata/firmware-auto-efi-format-nvram-raw-loader-path.xml copy tests/qemuxmlconfdata/{firmware-auto-efi-format-mismatch.x86_64-latest.args => firmware-auto-efi-format-nvram-raw-nvramtemplate-path.x86_64-latest.args} (90%) copy tests/qemuxmlconfdata/{firmware-auto-efi-format-mismatch.x86_64-latest.xml => firmware-auto-efi-format-nvram-raw-nvramtemplate-path.x86_64-latest.xml} (81%) create mode 100644 tests/qemuxmlconfdata/firmware-auto-efi-format-nvram-raw-nvramtemplate-path.xml copy tests/qemuxmlconfdata/{firmware-auto-efi-format-mismatch.x86_64-latest.args => firmware-auto-efi-format-nvram-raw.x86_64-latest.abi-update.args} (100%) copy tests/qemuxmlconfdata/{firmware-auto-efi-format-mismatch.x86_64-latest.xml => firmware-auto-efi-format-nvram-raw.x86_64-latest.abi-update.xml} (100%) copy tests/qemuxmlconfdata/{firmware-auto-efi-format-mismatch.x86_64-latest.args => firmware-auto-efi-format-nvram-raw.x86_64-latest.args} (100%) copy tests/qemuxmlconfdata/{firmware-auto-efi-format-mismatch.x86_64-latest.xml => firmware-auto-efi-format-nvram-raw.x86_64-latest.xml} (100%) create mode 100644 tests/qemuxmlconfdata/firmware-auto-efi-format-nvram-raw.xml copy tests/qemuxmlconfdata/{firmware-auto-efi-format-mismatch.x86_64-latest.args => firmware-auto-efi-format-nvramtemplate-qcow2.x86_64-latest.args} (77%) copy tests/qemuxmlconfdata/{firmware-auto-efi-format-mismatch.x86_64-latest.xml => firmware-auto-efi-format-nvramtemplate-qcow2.x86_64-latest.xml} (81%) create mode 100644 tests/qemuxmlconfdata/firmware-auto-efi-format-nvramtemplate-qcow2.xml create mode 100644 tests/qemuxmlconfdata/firmware-manual-bios-rw.x86_64-latest.err create mode 100644 tests/qemuxmlconfdata/firmware-manual-bios-rw.xml copy tests/qemuxmlconfdata/{firmware-auto-efi-format-mismatch.x86_64-latest.args => firmware-manual-efi-nvram-template-nonstandard-format.x86_64-latest.args} (70%) copy tests/qemuxmlconfdata/{firmware-manual-efi-nvram-template-nonstandard.x86_64-latest.xml => firmware-manual-efi-nvram-template-nonstandard-format.x86_64-latest.xml} (81%) create mode 100644 tests/qemuxmlconfdata/firmware-manual-efi-nvram-template-nonstandard-format.xml copy tests/qemuxmlconfdata/{firmware-auto-efi-format-mismatch.x86_64-latest.args => firmware-manual-efi-nvram-template-nonstandard-legacy-paths.x86_64-latest.args} (89%) copy tests/qemuxmlconfdata/{firmware-manual-efi-loader-path-nonstandard.x86_64-latest.xml => firmware-manual-efi-nvram-template-nonstandard-legacy-paths.x86_64-latest.xml} (81%) create mode 100644 tests/qemuxmlconfdata/firmware-manual-efi-nvram-template-nonstandard-legacy-paths.xml create mode 100644 tests/qemuxmlconfdata/firmware-manual-efi-rw-nvram.x86_64-latest.err create mode 100644 tests/qemuxmlconfdata/firmware-manual-efi-rw-nvram.xml copy tests/qemuxmlconfdata/{firmware-auto-efi-format-mismatch.x86_64-latest.args => firmware-manual-efi-sev-snp.x86_64-latest+amdsev.args} (74%) copy tests/qemuxmlconfdata/{firmware-manual-efi-nvram-template-nonstandard.x86_64-latest.xml => firmware-manual-efi-sev-snp.x86_64-latest+amdsev.xml} (75%) create mode 100644 tests/qemuxmlconfdata/firmware-manual-efi-sev-snp.xml copy tests/qemuxmlconfdata/{firmware-auto-efi-format-mismatch.x86_64-latest.args => firmware-manual-efi-tdx.x86_64-latest+inteltdx.args} (59%) copy tests/qemuxmlconfdata/{firmware-auto-efi-format-mismatch.x86_64-latest.xml => firmware-manual-efi-tdx.x86_64-latest+inteltdx.xml} (69%) create mode 100644 tests/qemuxmlconfdata/firmware-manual-efi-tdx.xml -- 2.52.0

3 38

[PATCH] tools: check for failure decoding secret base64 value
by Daniel P. Berrangé 06 Jan '26

06 Jan '26

From: Daniel P. Berrangé <berrange(a)redhat.com> The g_base64_decode function will return a valid pointer, but with length of zero when it fails to decode data. Report an error in that scenario, so avoid a later more obsecure error. eg old behaviour # virsh secret-set-value f52a81b2-424e-490c-823d-6bd4235bc507 foo warning: Passing secret value as command-line argument is insecure! error: Failed to set secret value error: value in virSecretSetValue must not be NULL new behaviour # /home/berrange/virsh secret-set-value f52a81b2-424e-490c-823d-6bd4235bc507 foo warning: Passing secret value as command-line argument is insecure! error: Secret value is not valid base64 Signed-off-by: Daniel P. Berrangé <berrange(a)redhat.com> --- tools/virsh-secret.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/tools/virsh-secret.c b/tools/virsh-secret.c index 26ecb41073..215e8e1017 100644 --- a/tools/virsh-secret.c +++ b/tools/virsh-secret.c @@ -267,6 +267,11 @@ cmdSecretSetValue(vshControl *ctl, const vshCmd *cmd) secret_val = (char *) g_base64_decode(tmp, &secret_len); virSecureErase(tmp, tmp_len); + + if (!secret_len) { + vshError(ctl, "%s", _("Secret value is not valid base64")); + return false; + } } res = virSecretSetValue(secret, (unsigned char *) secret_val, secret_len, 0); -- 2.52.0

2 1