- Devel - libvirt List Archives

[libvirt PATCH 00/16] qemu: implement virtiofs hot(un)plug (virtiofs epopee)
by Ján Tomko 06 Oct '21

06 Oct '21

https://bugzilla.redhat.com/show_bug.cgi?id=1897708 Ján Tomko (16): qemu: vhost-user-fs: format alias on the command line conf: define cleanup func for virDomainChrSourceDef qemu: domain: introduce qemuDomainGetVHostUserFSSocketPath qemu: domain: introduce qemuDomainGetVHostUserChrSourceDef tests: qemuxml2argvtest: fix path to virtiofs socket qemu: vhost-user-fs: separate building of chardev string qemu: vhost-user-fs: separate building of device string qemu: do not put virtiofs socket in private data qemu: remove private data from virDomainFSDef qemu: alias: prepare qemuAssignDeviceFSAlias for disjunct ranges qemu: vhost-user-fs: build extdevice for zpci qemu: export vhost-user-fs-related functions qemu: store logCtxt in domain private data qemu: use priv->logCtxt in qemuProcessLaunch qemu: implement virtiofs hotplug qemu: implement virtiofs hotunplug src/conf/domain_conf.c | 24 +++ src/conf/domain_conf.h | 4 + src/libvirt_private.syms | 1 + src/qemu/qemu_alias.c | 20 +- src/qemu/qemu_alias.h | 4 + src/qemu/qemu_command.c | 81 +++++--- src/qemu/qemu_command.h | 6 + src/qemu/qemu_domain.c | 64 +++---- src/qemu/qemu_domain.h | 30 +-- src/qemu/qemu_driver.c | 9 +- src/qemu/qemu_extdevice.c | 6 +- src/qemu/qemu_hotplug.c | 177 +++++++++++++++++- src/qemu/qemu_hotplug.h | 4 + src/qemu/qemu_process.c | 21 ++- src/qemu/qemu_virtiofs.c | 14 +- ...vhost-user-fs-fd-memory.x86_64-latest.args | 4 +- ...vhost-user-fs-hugepages.x86_64-latest.args | 4 +- tests/qemuxml2argvtest.c | 12 -- 18 files changed, 360 insertions(+), 125 deletions(-) -- 2.31.1

2 37

[PATCH 0/8] virsh: Use g_autoptr() for public types
by Michal Privoznik 05 Oct '21

05 Oct '21

In this patchset I'm switching from virXXXFree to g_autoptr(). There are still some left, but very rare occurrence: libvirt.git $ git grep -o "vir[A-Z].*Free" -- tools/ | \ cut -d':' -f 2 | sort | uniq -c | sort -n And of course, after these some functions could use subsequent cleanup, e.g. because cleanup label collapsed to just return statement. But I leave those for future work. Michal Prívozník (8): virsh-util.h: Fix ordering of virshXXXFree functions virsh: Add wrapper for virInterfaceFree virsh: Add wrapper for virStoragePoolFree virsh: Add wrapper for virStorageVolFree virsh: Add wrapper for virNetworkFree virsh: Add wrapper for virNodeDeviceFree virsh: Add wrapper for virNWFilterFree virsh: Add wrapper for virStreamFree build-aux/syntax-check.mk | 4 +- tools/virsh-completer-interface.c | 3 +- tools/virsh-completer-network.c | 8 +-- tools/virsh-completer-nodedev.c | 3 +- tools/virsh-completer-nwfilter.c | 3 +- tools/virsh-completer-pool.c | 3 +- tools/virsh-completer-volume.c | 8 +-- tools/virsh-console.c | 8 +-- tools/virsh-domain.c | 10 +-- tools/virsh-interface.c | 55 +++++---------- tools/virsh-network.c | 75 ++++++-------------- tools/virsh-nodedev.c | 49 ++++---------- tools/virsh-nwfilter.c | 22 ++---- tools/virsh-pool.c | 67 ++++++------------ tools/virsh-util.c | 78 +++++++++++++++++++++ tools/virsh-util.h | 46 +++++++++++-- tools/virsh-volume.c | 109 ++++++++---------------------- 17 files changed, 254 insertions(+), 297 deletions(-) -- 2.32.0

2 9

[libvirt PATCH v3 0/9] Automatic mutex management
by Tim Wiederhake 05 Oct '21

05 Oct '21

V1: https://listman.redhat.com/archives/libvir-list/2021-August/msg00823.html V2: https://listman.redhat.com/archives/libvir-list/2021-September/msg00249.html Changes since V2: * Dropped VIR_XPATH_NODE_AUTORESTORE simplification. Moved to a separate series. * Dropped the attempt to work around g_auto* / clang unused variable warnings (See https://bugs.llvm.org/show_bug.cgi?id=3888, https://bugs.llvm.org/show_bug.cgi?id=43482, and https://gitlab.gnome.org/GNOME/glib/-/merge_requests/2272) Instead, introduce VIR_LOCK_GUARD macro that adds G_GNUC_UNUSED. Regards, Tim Tim Wiederhake (9): internal: Add CONCAT macro virthread: Introduce virLockGuard virthread: Introduce VIR_WITH_MUTEX_LOCK_GUARD virobject: Introduce virObjectLockGuard virobject: Introduce VIR_WITH_OBJECT_LOCK_GUARD virChrdevFDStreamCloseCb: Use virLockGuardNew virChrdevFree: Use VIR_WITH_MUTEX_LOCK bhyveAutostartDomain: Use virObjectLockGuard lxcDomainDetachDeviceHostdevUSBLive: Use VIR_WITH_OBJECT_LOCK_GUARD src/bhyve/bhyve_driver.c | 4 ++-- src/conf/virchrdev.c | 12 +++++------- src/internal.h | 3 +++ src/libvirt_private.syms | 4 ++++ src/lxc/lxc_driver.c | 6 +++--- src/util/virobject.c | 16 ++++++++++++++++ src/util/virobject.h | 24 ++++++++++++++++++++++++ src/util/virthread.c | 26 ++++++++++++++++++++++++++ src/util/virthread.h | 31 +++++++++++++++++++++++++++++++ 9 files changed, 114 insertions(+), 12 deletions(-) -- 2.31.1

2 10

[PATCH] qemu: Check if unpriv_sgio is already set before trying to set it
by Michal Privoznik 05 Oct '21

05 Oct '21

In case when libvirt runs inside a restricted container it may not have enough permissions to modify unpriv_sgio. However, it may have been set beforehand by sysadmin or an orchestration tool. Therefore, let's check whether the currently set value is the one we want and if it is refrain from writing to the file. Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=2010306 Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- src/qemu/qemu_conf.c | 14 +++++++++++--- 1 file changed, 11 insertions(+), 3 deletions(-) diff --git a/src/qemu/qemu_conf.c b/src/qemu/qemu_conf.c index 916a3d36ee..0451bc70ac 100644 --- a/src/qemu/qemu_conf.c +++ b/src/qemu/qemu_conf.c @@ -1878,9 +1878,17 @@ qemuSetUnprivSGIO(virDomainDeviceDef *dev) * whitelist is enabled. But if requesting unfiltered access, always call * virSetDeviceUnprivSGIO, to report an error for unsupported unpriv_sgio. */ - if ((virFileExists(sysfs_path) || val == 1) && - virSetDeviceUnprivSGIO(path, NULL, val) < 0) - return -1; + if (virFileExists(sysfs_path) || val == 1) { + int curr_val; + + if (virGetDeviceUnprivSGIO(path, NULL, &curr_val) < 0) + return -1; + + if (curr_val != val && + virSetDeviceUnprivSGIO(path, NULL, val) < 0) { + return -1; + } + } return 0; } -- 2.32.0

2 1

[PATCH] qemu: capabilities: rename piix4-acpi-root-hotplug-en to more appropriate name
by Ani Sinha 05 Oct '21

05 Oct '21

The capability name piix4-acpi-root-hotplug-en is not conventional and appreared to be confusing to some. "en" suffix is also incorrect as the capability in qemu is used to both enable and disable hotplug on the pci root bus on the i440fx. Hence, rename it to piix4.acpi-root-pci-hotplug so that it is clearer, less confusing and more accurate. Signed-off-by: Ani Sinha <ani(a)anisinha.ca> --- src/qemu/qemu_capabilities.c | 2 +- tests/qemucapabilitiesdata/caps_5.2.0.x86_64.xml | 2 +- tests/qemucapabilitiesdata/caps_6.0.0.x86_64.xml | 2 +- tests/qemucapabilitiesdata/caps_6.1.0.x86_64.xml | 2 +- 4 files changed, 4 insertions(+), 4 deletions(-) diff --git a/src/qemu/qemu_capabilities.c b/src/qemu/qemu_capabilities.c index 9d0c96a20c..d7b79ef7c0 100644 --- a/src/qemu/qemu_capabilities.c +++ b/src/qemu/qemu_capabilities.c @@ -641,7 +641,7 @@ VIR_ENUM_IMPL(virQEMUCaps, "virtio-blk.queue-size", /* QEMU_CAPS_VIRTIO_BLK_QUEUE_SIZE */ /* 410 */ - "piix4-acpi-root-hotplug-en", /* QEMU_CAPS_PIIX_ACPI_ROOT_PCI_HOTPLUG */ + "piix4.acpi-root-pci-hotplug", /* QEMU_CAPS_PIIX_ACPI_ROOT_PCI_HOTPLUG */ ); diff --git a/tests/qemucapabilitiesdata/caps_5.2.0.x86_64.xml b/tests/qemucapabilitiesdata/caps_5.2.0.x86_64.xml index 1e5833a9f0..834fb86636 100644 --- a/tests/qemucapabilitiesdata/caps_5.2.0.x86_64.xml +++ b/tests/qemucapabilitiesdata/caps_5.2.0.x86_64.xml @@ -230,7 +230,7 @@ <flag name='input-linux'/> <flag name='query-display-options'/> <flag name='virtio-blk.queue-size'/> - <flag name='piix4-acpi-root-hotplug-en'/> + <flag name='piix4.acpi-root-pci-hotplug'/> <version>5002000</version> <kvmVersion>0</kvmVersion> <microcodeVersion>43100243</microcodeVersion> diff --git a/tests/qemucapabilitiesdata/caps_6.0.0.x86_64.xml b/tests/qemucapabilitiesdata/caps_6.0.0.x86_64.xml index b54dd8a22e..e9d1a26400 100644 --- a/tests/qemucapabilitiesdata/caps_6.0.0.x86_64.xml +++ b/tests/qemucapabilitiesdata/caps_6.0.0.x86_64.xml @@ -238,7 +238,7 @@ <flag name='query-display-options'/> <flag name='set-action'/> <flag name='virtio-blk.queue-size'/> - <flag name='piix4-acpi-root-hotplug-en'/> + <flag name='piix4.acpi-root-pci-hotplug'/> <version>6000000</version> <kvmVersion>0</kvmVersion> <microcodeVersion>43100242</microcodeVersion> diff --git a/tests/qemucapabilitiesdata/caps_6.1.0.x86_64.xml b/tests/qemucapabilitiesdata/caps_6.1.0.x86_64.xml index 0ad493191d..971d55e0cc 100644 --- a/tests/qemucapabilitiesdata/caps_6.1.0.x86_64.xml +++ b/tests/qemucapabilitiesdata/caps_6.1.0.x86_64.xml @@ -240,7 +240,7 @@ <flag name='query-display-options'/> <flag name='set-action'/> <flag name='virtio-blk.queue-size'/> - <flag name='piix4-acpi-root-hotplug-en'/> + <flag name='piix4.acpi-root-pci-hotplug'/> <version>6001000</version> <kvmVersion>0</kvmVersion> <microcodeVersion>43100243</microcodeVersion> -- 2.25.1

2 2

[PATCH 0/5] ch: Fixup object leaks and extend g_auto usage
by William Douglas 05 Oct '21

05 Oct '21

Based off of feedback from Laine to Ján's series of patches increasing g_auto usage for the ch_driver. This fixes leaks in virCHMonitorBuildKernelRelatedJson and virCHMonitorNew while tidying up and increasing the automatic cleanup for those functions and virCHMonitorBuildMemoryJson. William Douglas (5): ch_monitor: Stop leaking json value objects ch_monitor: Correctly close and ref the virCHMonitor ch: use g_auto in virCHMonitorBuildMemoryJson ch: use g_auto in virCHMonitorBuildKernelRelatedJson ch: use g_auto in virCHMonitorNew src/ch/ch_monitor.c | 59 +++++++++++++++------------------------------ src/ch/ch_monitor.h | 1 + 2 files changed, 20 insertions(+), 40 deletions(-) -- 2.33.0

3 11

[PATCH v7 0/4] Add support to enable/disable hotplug on pci-root controller
by Ani Sinha 04 Oct '21

04 Oct '21

changelog: v7: rebased the patches post libvirt release 7.8. Adjusted NEWS.rst to reflect 7.9 (unreleased) version. Changed version info for new config option in formatdomain.rst as well. Added reviewed-by tags. Also in formatdomain.rst added information on what type of hotplug (ACPI/native etc) are affected by the setting. v6: incorporated Jan's suggestions. v5: incorporated suggestions from Laine for patches #2 and #3. The qemu command line is now added in a new function and called from qemuBuildControllersByTypeCommandLine(). Output xmls are now symlinked to input xmls for unit tests. v4: split the original patchset into a pci-root controller specific patch series. also the libvirt conf is now a sub-element of the pci-root controller as was suggested by Dan Berrange. Please see discussion here: https://listman.redhat.com/archives/libvir-list/2021-September/msg00839.html v3: reorganized the patches as per Laine's suggestion. Added more details in commit messages. Added conf description in formatdomain.rst. Added changelog for next release. v2: fixed bugs and added additional missing unit tests. v1: initial implementation. Had some bugs and missed some unit tests This patchset introduces libvirt xml support to enable/disable hotplug on the pci-root controller. It adds a 'target' subelement for the pci-root controller with a 'hotplug' property. This property can be used to enable or disable hotplug for the pci-root controller. For example, in order to disable hotplug on the pci-root controller, one has to use set '<target hotplug='off'>' as shown below: <controller type='pci' model='pci-root'> <target hotplug='off'/> </controller> '<target hotplug='on'>' option would enable hotplug for pci-root controller. This is also the default value. This option is only available for pc machine types and is applicable for qemu only/kvm accelerator onlt.This feature was introduced from qemu version 5.2 with the following change in qemu repository: 3d7e78aa7777f ("Introduce a new flag for i440fx to disable PCI hotplug on the root bus") The above qemu commit describes some reasons why users might to disable hotplug on PCI root buses [1]. The corresponding commandline option to qemu for x86 guests is: -global PIIX4_PM.acpi-root-pci-hotplug=<off/on> Notes: 1. The use case scenario described by Laine in https://listman.redhat.com/archives/libvir-list/2020-February/msg00110.html intentionally does not discuss i440fx and focusses solely on q35. I do realize that redhat has moved on from i440fx and currently efforts for new features are concentrated on q35 machines only. We have had some hard debates on this on the qemu mailing list before. The fact of the matter is that i440fx is not at 1-1 parity with q35. There are many users who are currenly using i440fx and are simply not ready to move to q35 without sacrificing some existing features they support today. For example https://wiki.qemu.org/images/4/4e/Q35.pdf lists some of q35 limitations. https://www.linux-kvm.org/images/0/06/2012-forum-Q35.pdf provides more information on the differences. Hence we need to solve the issue Laine has described in the above email for i440fx without adding additional bridges. Further, in Daniel Berrange's words from : https://lists.gnu.org/archive/html/qemu-devel/2020-04/msg03012.html "From the upstream POV, there's been no decision / agreement to phase out PIIX, this is purely a RHEL downstream decision & plan. If other distros / users have a different POV, and find the feature useful, we should accept the patch if it meets the normal QEMU patch requirements. " Also to be noted that I have already experimented this qemu commandline option using libvirt passthrough feature as has been documented in http://blog.vmsplice.net/2011/04/how-to-pass-qemu-command-line-options.html This was only meant to be a short term solution until libvirt started supporting this natively. Supporting this option through libvirt would simplify their use case as well as add capability validations and graceful failure scenarios in case qemu did not support the option. Ani Sinha (4): qemu: capablities: detect presence of acpi-root-pci-hotplug for i440fx machines conf: introduce option to enable/disable pci hotplug on pci-root controller qemu: command: add support to enable/disable hotplug on pci-root controller NEWS: release note the new hotplug enable/disable option on pci-root controller NEWS.rst | 6 ++++ docs/formatdomain.rst | 14 ++++++--- src/qemu/qemu_capabilities.c | 4 +++ src/qemu/qemu_capabilities.h | 3 ++ src/qemu/qemu_command.c | 17 ++++++++++ src/qemu/qemu_validate.c | 9 +++++- .../caps_5.2.0.x86_64.xml | 1 + .../caps_6.0.0.x86_64.xml | 1 + .../caps_6.1.0.x86_64.xml | 1 + .../pc-i440fx-acpi-root-hotplug-disable.args | 31 +++++++++++++++++++ .../pc-i440fx-acpi-root-hotplug-disable.err | 1 + .../pc-i440fx-acpi-root-hotplug-disable.xml | 30 ++++++++++++++++++ .../pc-i440fx-acpi-root-hotplug-enable.xml | 30 ++++++++++++++++++ tests/qemuxml2argvtest.c | 3 ++ .../pc-i440fx-acpi-root-hotplug-disable.xml | 1 + .../pc-i440fx-acpi-root-hotplug-enable.xml | 1 + tests/qemuxml2xmltest.c | 4 +++ 17 files changed, 151 insertions(+), 6 deletions(-) create mode 100644 tests/qemuxml2argvdata/pc-i440fx-acpi-root-hotplug-disable.args create mode 100644 tests/qemuxml2argvdata/pc-i440fx-acpi-root-hotplug-disable.err create mode 100644 tests/qemuxml2argvdata/pc-i440fx-acpi-root-hotplug-disable.xml create mode 100644 tests/qemuxml2argvdata/pc-i440fx-acpi-root-hotplug-enable.xml create mode 120000 tests/qemuxml2xmloutdata/pc-i440fx-acpi-root-hotplug-disable.xml create mode 120000 tests/qemuxml2xmloutdata/pc-i440fx-acpi-root-hotplug-enable.xml -- 2.25.1

3 11

[PATCH] NEWS: cosmetic - fix indentation
by Ani Sinha 04 Oct '21

04 Oct '21

The indentation of the first item under the categoty "new features" for the future release v7.9.0 is not right. Fix it. Signed-off-by: Ani Sinha <ani(a)anisinha.ca> --- NEWS.rst | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/NEWS.rst b/NEWS.rst index 76d2375e97..caa61f0646 100644 --- a/NEWS.rst +++ b/NEWS.rst @@ -17,7 +17,7 @@ v7.9.0 (unreleased) * **New features** - * Introduce virtio-mem ``<memory/>`` model + * Introduce virtio-mem ``<memory/>`` model New virtio-mem model is introduced for ``<memory/>`` device which is a paravirtualized mechanism of adding/removing memory to/from a VM. Use -- 2.25.1

2 2

[PATCH] virsh: Fix --nvram and --keep-nvram help strings
by Michal Privoznik 04 Oct '21

04 Oct '21

The --nvram and --keep-nvram options of the undefine command can be used regardless of the domain status (the only consumer so far - qemuDomainUndefineFlags() doesn't care about the domain status). Yet, their corresponding help strings say something about inactive domains while manpage says nothing. Remove the reference to domain state. Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=2007659 Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- tools/virsh-domain.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/virsh-domain.c b/tools/virsh-domain.c index 0b78fbf728..f73d9a057e 100644 --- a/tools/virsh-domain.c +++ b/tools/virsh-domain.c @@ -3595,11 +3595,11 @@ static const vshCmdOptDef opts_undefine[] = { }, {.name = "nvram", .type = VSH_OT_BOOL, - .help = N_("remove nvram file, if inactive") + .help = N_("remove nvram file") }, {.name = "keep-nvram", .type = VSH_OT_BOOL, - .help = N_("keep nvram file, if inactive") + .help = N_("keep nvram file") }, {.name = NULL} }; -- 2.32.0

2 1

[PATCH 0/1] vmx: Fix <genid/> mapping
by Michal Privoznik 04 Oct '21

04 Oct '21

Apparently, parsing vmx.genid is not as easy as I thought. Anyway, it was brought up in a private thread that libvirt doesn't report correct UUIDs. For instance for the following input: vm.genid = "-8536691797830587195" vm.genidX = "-1723453263670062919" Libvirt would report: <genid>8987940a-0951-2cc5-e815-10634ff550b9</genid> while virt-v2v would report: <genid>09512cc5-940a-8987-b950-f54f631015e8</genid> Another example: vm.genid = "-6284418052148993891" vm.genidX = "-649955058627554545" Libvirt: <genid>a8c94313-ed9b-609d-f6fa-e5515a89530f</genid> virt-v2v: <genid>ed9b609d-4313-a8c9-0f53-895a51e5faf6</genid> To test my patch I modified tests/vmx2xmldata/esx-in-the-wild-10.vmx (because it already contains vmx.genid, but any file can be modified really), and then ran: libvirt.git/_build/tests $ VIR_TEST_DEBUG=1 VIR_TEST_RANGE=58 ./vmx2xmltest Mind you, the fix is almost direct rewrite of virt-v2v's algorithm, except it's optimized for C and not OCaml :-) You can find various attempts/versions of that on my gitlab: https://gitlab.com/MichalPrivoznik/libvirt/-/commits/vmx_genid/ I'm sending only the last one because that looks the best IMO. Michal Prívozník (1): vmx: Fix <genid/> mapping src/vmx/vmx.c | 16 ++++++++++++---- tests/vmx2xmldata/esx-in-the-wild-10.xml | 2 +- 2 files changed, 13 insertions(+), 5 deletions(-) -- 2.32.0

4 17