April 2020 - Devel - libvirt List Archives

[PATCH v4 0/7] Tighten qemu-img rules on missing backing format
by Eric Blake 03 Apr '20

03 Apr '20

3 14

Release of libvirt-6.2.0
by Daniel Veillard 03 Apr '20

03 Apr '20

Mea-culpa I forgot to push RC2 on Tuesday, rather than pushing the release off by 2 more days, I decided to push the release today to stay on schedule. As a result 6.2.0 is out, tagged in git and source tarball and rpm are available at the usual place: https://libvirt.org/sources/ I also pushed a release of the python binding with minimal changes that can be found at: https://libvirt.org/sources/libvirt/ New features: - qemu: NVDIMM support for pSeries guests QEMU 5.0 implements NVDIMM memory support for pSeries guests. This is done by adding an 'uuid' element in the memory XML, which can either be provided in the XML or, if omitted, generated automatically. - qemu: Add virtiofs support This feature, introduced in QEMU 4.2, is a more modern alternative to virtio-9p, which is exposed through the same <filesystem/> element. - admin: Support reloading TLS certificates After renewing TLS certificates, it was usually necessary to restart libvirtd for the new ones to be loaded: now the same result can be obtained without restarting the daemon by using virt-admin server-update-tls. Removed features: - Removed support for INI style of comments With switching of our internal code to GLib, parsing of client authentication config files is handed over to GLib which does not support INI style of comments starting with a semicolon (;). Use number sign (#) instead. * Improvements - qemu: Don't compare local and remote hostnames on migration This check was introduced to prevent same-host migration, but did not work as expected when multiple libvirtd instances were running on the same host but in different containers. With this release, the host UUID (which should be unique to the container) is checked instead. - qemu: Use per-VM event loops Instead of using a single even loop to process communication with the QEMU monitor and guest agent, create a separate one for each VM. This helps with scalability and prevents scenarios where a single malfunctioning VM could affect all those running on the same host. - qemu: Support migration with SLIRP helper interface With QEMU 5.0, a new D-Bus backend allows migration of external processes. When needed, libvirt will start a per-vm D-Bus bus, and migrate the slirp-helper along with QEMU. Bug fixes: - qemu: Open backing chain late for shallow block copy reusing external images With introduction of -blockdev for QEMU storage configuration in libvirt-5.10 we've started opening the backing chain of the destination/mirror of a virDomainBlockcopy started with VIR_DOMAIN_BLOCK_COPY_REUSE_EXT | VIR_DOMAIN_BLOCK_COPY_SHALLOW flags when starting the job rather than when virDomainBlockJobAbort with VIR_DOMAIN_BLOCK_JOB_ABORT_PIVOT is issued. For users depending on this undocumented quirky pre-blockdev behaviour this caused a regression as the backing chain could not be modified while the copy of the top image was progressing due to QEMU image locking. Note that this fix also requires qemu-5.0 while -blockdev is used starting from QEMU-4.2. - Don't generate machine names containing dots Even though the guest name containing dots is not a problem for libvirt itself, we need to strip them out when registering with machined because of the latter's requirements. Thanks everybody who contributed to this release be it with bug reports, fixes, reviews, documentation, etc ... Enjoy the release, stay safe! Daniel -- Daniel Veillard | Red Hat Developers Tools http://developer.redhat.com/ veillard(a)redhat.com | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ http://veillard.com/ | virtualization library http://libvirt.org/

1 0

SRPMs as download artifacts (was: Re: libvirt Source RPMs for CentOS or RHEL?)
by Andrea Bolognani 03 Apr '20

03 Apr '20

[moving to libvir-list] On Thu, 2020-04-02 at 09:41 +0100, Daniel P. Berrangé wrote: > On Thu, Apr 02, 2020 at 08:47:30AM +0800, FuLong Wang wrote: > > Do we have libvirt source rpms (version above 5.9.0) for CentOS or RHEL? > > > > I only find source rpms for fedora in below public link. > > > > https://libvirt.org/sources/ > > Please ignore all the RPMs you see there, they really shouldn't be > used. Can we just stop generating them? And get rid of all the existing ones, especially silly stuff like the binary RPMs from 2013 targeting Fedora 17. This will have the additional benefit of making it easier to find what you're looking for, since it will no longer be hidden among a heap of irrelevant garbage. > The source tarballs (eg libvirt-5.9.0.tar.xz) contain a spec file > inside. > > This means you can generate RPMs for your precise distro using something > akin to the following commands: > > $ rpmbuild -ts libvirt-5.9.0.tar.xz > $ sudo dnf install redhat-rpm-config > $ sudo dnf builddep $HOME/rpmbuild/SRPMS/libvirt-5.9.0-1.fc31.src.rpm > $ rpmbuild --rebuild $HOME/rpmbuild/SRPMS/libvirt-5.9.0-1.fc31.src.rpm I think it would make sense to document this quick procedure somewhere, if we haven't already. There's some obvious overlap with our 'make rpm' target, but that one requires unpacking the release archive and running configure beforehand, so it's not quite as nice for someone who just wants to quickly get a working set of RPMs. -- Andrea Bolognani / Red Hat / Virtualization

2 2

disable passwd access to spice and specify disable-ticketing
by Vasiliy Tolstov 03 Apr '20

03 Apr '20

Hi! i need to have passwordless auth to spice (because it is restricted to 127.0.0.1 and nobody can connect to it, in feature i want to use it via unix socket). I'm checking docs at libvirt site and don't see any disable-ticketing options. SO how can I pass this option and allow access without a password to spice? Thanks! -- Vasiliy Tolstov, e-mail: v.tolstov(a)selfip.ru

2 2

request: wiki account
by Sebastian Mitterle 03 Apr '20

03 Apr '20

As mentioned on https://wiki.libvirt.org/page/Main_Page I'd like to request a wiki account for contributions on wiki.libvirt.org Preferred username: smitterl Thank you in advance, Sebastian

1 0

[libvirt PATCH 0/8] daemons: Improve timeout handling, plus some cleanups
by Andrea Bolognani 03 Apr '20

03 Apr '20

The main motivation for this series is explained in patch 2/8, but of course in the process I managed to locate a number of yaks that were looking quite hairy indeed... Andrea Bolognani (8): daemons: Support --timeout 0 Revert "remote: move timeout arg into sysconf file" Revert "logging: Use default timeout of 120 seconds for virtlogd" logging, locking: Set default timeout of 120 seconds Add sysconf files for all daemons spec: Include new sysconf files qemu: Allow audio driver override in virtqemud remote: Drop KRB5_KTNAME override libvirt.spec.in | 11 +++++++++++ src/interface/Makefile.inc.am | 2 ++ src/interface/virtinterfaced.service.in | 3 ++- src/interface/virtinterfaced.sysconf | 3 +++ src/libxl/Makefile.inc.am | 2 ++ src/libxl/virtxend.service.in | 3 ++- src/libxl/virtxend.sysconf | 3 +++ src/locking/lock_daemon.c | 6 +++--- src/locking/virtlockd.service.in | 2 +- src/logging/log_daemon.c | 6 +++--- src/logging/virtlogd.service.in | 2 +- src/logging/virtlogd.sysconf | 2 +- src/lxc/Makefile.inc.am | 2 ++ src/lxc/virtlxcd.service.in | 3 ++- src/lxc/virtlxcd.sysconf | 3 +++ src/network/Makefile.inc.am | 2 ++ src/network/virtnetworkd.service.in | 3 ++- src/network/virtnetworkd.sysconf | 3 +++ src/node_device/Makefile.inc.am | 2 ++ src/node_device/virtnodedevd.service.in | 3 ++- src/node_device/virtnodedevd.sysconf | 3 +++ src/nwfilter/Makefile.inc.am | 2 ++ src/nwfilter/virtnwfilterd.service.in | 3 ++- src/nwfilter/virtnwfilterd.sysconf | 3 +++ src/qemu/Makefile.inc.am | 2 ++ src/qemu/virtqemud.service.in | 3 ++- src/qemu/virtqemud.sysconf | 12 ++++++++++++ src/remote/Makefile.inc.am | 5 ++++- src/remote/libvirtd.sasl | 4 +--- src/remote/libvirtd.service.in | 6 +++++- src/remote/libvirtd.sysconf | 15 +++------------ src/remote/remote_daemon.c | 6 +++--- src/remote/virtproxyd.service.in | 3 ++- src/remote/virtproxyd.sysconf | 3 +++ src/secret/Makefile.inc.am | 2 ++ src/secret/virtsecretd.service.in | 3 ++- src/secret/virtsecretd.sysconf | 3 +++ src/storage/Makefile.inc.am | 2 ++ src/storage/virtstoraged.service.in | 3 ++- src/storage/virtstoraged.sysconf | 3 +++ src/vbox/Makefile.inc.am | 2 ++ src/vbox/virtvboxd.service.in | 3 ++- src/vbox/virtvboxd.sysconf | 3 +++ src/vz/Makefile.inc.am | 2 ++ src/vz/virtvzd.service.in | 3 ++- src/vz/virtvzd.sysconf | 3 +++ 46 files changed, 127 insertions(+), 41 deletions(-) create mode 100644 src/interface/virtinterfaced.sysconf create mode 100644 src/libxl/virtxend.sysconf create mode 100644 src/lxc/virtlxcd.sysconf create mode 100644 src/network/virtnetworkd.sysconf create mode 100644 src/node_device/virtnodedevd.sysconf create mode 100644 src/nwfilter/virtnwfilterd.sysconf create mode 100644 src/qemu/virtqemud.sysconf create mode 100644 src/remote/virtproxyd.sysconf create mode 100644 src/secret/virtsecretd.sysconf create mode 100644 src/storage/virtstoraged.sysconf create mode 100644 src/vbox/virtvboxd.sysconf create mode 100644 src/vz/virtvzd.sysconf -- 2.25.1

4 32

[PATCH 0/2] Optimize initialization of storage files
by Peter Krempa 03 Apr '20

03 Apr '20

See 2/2. Peter Krempa (2): qemuSecurityChownCallback: Remove 'cleanup' section qemuSecurityChownCallback: Don't initialize storage file subsystem for local file src/qemu/qemu_driver.c | 31 +++++++++++++------------------ 1 file changed, 13 insertions(+), 18 deletions(-) -- 2.24.1

2 4

[PATCH 0/5] qemu: caps: Bump capabilities to qemu-5.0.0-rc1
by Peter Krempa 02 Apr '20

02 Apr '20

Peter Krempa (5): qemumonitorjsontest: Space out test name concatenation qemuMonitorTestProcessCommandDefaultValidate: Output validator output to stderr qemumonitorjsontest: AddNetdev: Use real variant syntax-check: Exempt tests/qemucapabilitiesdata from 'sc_prohibit_devname' qemu: capabilities: Update qemu-5.0.0 capabilities for x86_64 to rc1 build-aux/syntax-check.mk | 2 +- .../caps_5.0.0.x86_64.replies | 2314 +++++++++++------ .../caps_5.0.0.x86_64.xml | 144 +- tests/qemumonitorjsontest.c | 6 +- tests/qemumonitortestutils.c | 13 +- 5 files changed, 1707 insertions(+), 772 deletions(-) -- 2.25.1

2 6

[PATCH 0/3] Fix use-after-free in block job reconnection code
by Peter Krempa 02 Apr '20

02 Apr '20

Peter Krempa (3): qemu: migration: Don't use return value of qemuBlockJobUpdate qemuBlockJobUpdate: Remove return value qemuBlockJobRefreshJobs: Warn readers that 'job' may be invalid after update src/qemu/qemu_blockjob.c | 9 +++------ src/qemu/qemu_blockjob.h | 7 ++++--- src/qemu/qemu_migration.c | 17 +++++++---------- 3 files changed, 14 insertions(+), 19 deletions(-) -- 2.24.1

3 7

[PATCH for 6.2.0] qemuDomainSnapshotDiskPrepareOne: Fix logic of relative backing store update
by Peter Krempa 02 Apr '20

02 Apr '20

Commit 2ace7a87a8aced68c250 introduced a logic bug by an improperly modified condition where we'd skip to the else branch when reusing of external images was requested and blockdev is available. The original intentions were to skip the backing store update with blockdev. Fix it by only asserting the boolean which was used to track whether we support update of the backing store only when blockdev is not present along with the appropriate rename. https://bugzilla.redhat.com/show_bug.cgi?id=1820016 Signed-off-by: Peter Krempa <pkrempa(a)redhat.com> --- src/qemu/qemu_driver.c | 19 +++++++++++-------- 1 file changed, 11 insertions(+), 8 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 78024614cf..ff97f10f11 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -15129,7 +15129,7 @@ qemuDomainSnapshotDiskPrepareOne(virQEMUDriverPtr driver, { virDomainDiskDefPtr persistdisk; bool supportsCreate; - bool supportsBacking; + bool updateRelativeBacking = false; dd->disk = disk; @@ -15158,19 +15158,22 @@ qemuDomainSnapshotDiskPrepareOne(virQEMUDriverPtr driver, } supportsCreate = virStorageFileSupportsCreate(dd->src); - supportsBacking = virStorageFileSupportsBackingChainTraversal(dd->src); - if (supportsCreate || supportsBacking) { + /* relative backing store paths need to be updated so that relative + * block commit still works. With blockdev we must update it when doing + * commit anyways so it's skipped here */ + if (!blockdev && + virStorageFileSupportsBackingChainTraversal(dd->src)) + updateRelativeBacking = true; + + if (supportsCreate || updateRelativeBacking) { if (qemuDomainStorageFileInit(driver, vm, dd->src, NULL) < 0) return -1; dd->initialized = true; - /* relative backing store paths need to be updated so that relative - * block commit still works. With blockdev we must update it when doing - * commit anyways so it's skipped here */ - if (reuse && !blockdev) { - if (supportsBacking) { + if (reuse) { + if (updateRelativeBacking) { g_autofree char *backingStoreStr = NULL; if (virStorageFileGetBackingStoreStr(dd->src, &backingStoreStr) < 0) -- 2.25.1

2 1