April 2017 - Devel - libvirt List Archives

[libvirt] [PATCH] Fix python error reporting for some storage operations
by Cole Robinson 07 Mar '25

07 Mar '25

In the python bindings, all vir* classes expect to be passed a virConnect object when instantiated. Before the storage stuff, these classes were only instantiated in virConnect methods, so the generator is hardcoded to pass 'self' as the connection instance to these classes. Problem is there are some methods that return pool or vol instances which aren't called from virConnect: you can lookup a storage volume's associated pool, and can lookup volumes from a pool. In these cases passing 'self' doesn't give the vir* instance a connection, so when it comes time to raise an exception crap hits the fan. Rather than rework the generator to accomodate this edge case, I just fixed the init functions for virStorage* to pull the associated connection out of the passed value if it's not a virConnect instance. Thanks, Cole diff --git a/python/generator.py b/python/generator.py index 01a17da..c706b19 100755 --- a/python/generator.py +++ b/python/generator.py @@ -962,8 +962,12 @@ def buildWrappers(): list = reference_keepers[classname] for ref in list: classes.write(" self.%s = None\n" % ref[1]) - if classname in [ "virDomain", "virNetwork", "virStoragePool", "virStorageVol" ]: + if classname in [ "virDomain", "virNetwork" ]: classes.write(" self._conn = conn\n") + elif classname in [ "virStorageVol", "virStoragePool" ]: + classes.write(" self._conn = conn\n" + \ + " if not isinstance(conn, virConnect):\n" + \ + " self._conn = conn._conn\n") classes.write(" if _obj != None:self._o = _obj;return\n") classes.write(" self._o = None\n\n"); destruct=None

4 3

[libvirt] [PATCH] Fix compile error for stable 1.2.9
by Yang hongyang 27 Oct '23

27 Oct '23

Seems a backport miss. An extra member is passed to struct virLXCBasicMountInfo. Signed-off-by: Yang hongyang <hongyang.yang(a)easystack.cn> --- src/lxc/lxc_container.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/src/lxc/lxc_container.c b/src/lxc/lxc_container.c index 28dabec..1c65fa9 100644 --- a/src/lxc/lxc_container.c +++ b/src/lxc/lxc_container.c @@ -760,7 +760,7 @@ typedef struct { static const virLXCBasicMountInfo lxcBasicMounts[] = { { "proc", "/proc", "proc", MS_NOSUID|MS_NOEXEC|MS_NODEV, false, false }, - { "/proc/sys", "/proc/sys", NULL, MS_BIND|MS_NOSUID|MS_NOEXEC|MS_NODEV|MS_RDONLY, false, false, false }, + { "/proc/sys", "/proc/sys", NULL, MS_BIND|MS_NOSUID|MS_NOEXEC|MS_NODEV|MS_RDONLY, false, false }, { "sysfs", "/sys", "sysfs", MS_NOSUID|MS_NOEXEC|MS_NODEV|MS_RDONLY, false, false }, { "securityfs", "/sys/kernel/security", "securityfs", MS_NOSUID|MS_NOEXEC|MS_NODEV|MS_RDONLY, true, true }, #if WITH_SELINUX -- 1.7.1

3 2

[libvirt] [PATCH] web docs -- macvtap mode explanation
by Stefan Berger 27 Oct '23

27 Oct '23

This adds more information about the different macvtap device modes. Regards, Stefan

3 2

[libvirt] Supporting vhost-net and macvtap in libvirt for QEMU
by Anthony Liguori 27 Oct '23

27 Oct '23

Disclaimer: I am neither an SR-IOV nor a vhost-net expert, but I've CC'd people that are who can throw tomatoes at me for getting bits wrong :-) I wanted to start a discussion about supporting vhost-net in libvirt. vhost-net has not yet been merged into qemu but I expect it will be soon so it's a good time to start this discussion. There are two modes worth supporting for vhost-net in libvirt. The first mode is where vhost-net backs to a tun/tap device. This is behaves in very much the same way that -net tap behaves in qemu today. Basically, the difference is that the virtio backend is in the kernel instead of in qemu so there should be some performance improvement. Current, libvirt invokes qemu with -net tap,fd=X where X is an already open fd to a tun/tap device. I suspect that after we merge vhost-net, libvirt could support vhost-net in this mode by just doing -net vhost,fd=X. I think the only real question for libvirt is whether to provide a user visible switch to use vhost or to just always use vhost when it's available and it makes sense. Personally, I think the later makes sense. The more interesting invocation of vhost-net though is one where the vhost-net device backs directly to a physical network card. In this mode, vhost should get considerably better performance than the current implementation. I don't know the syntax yet, but I think it's reasonable to assume that it will look something like -net tap,dev=eth0. The effect will be that eth0 is dedicated to the guest. On most modern systems, there is a small number of network devices so this model is not all that useful except when dealing with SR-IOV adapters. In that case, each physical device can be exposed as many virtual devices (VFs). There are a few restrictions here though. The biggest is that currently, you can only change the number of VFs by reloading a kernel module so it's really a parameter that must be set at startup time. I think there are a few ways libvirt could support vhost-net in this second mode. The simplest would be to introduce a new tag similar to <source network='br0'>. In fact, if you probed the device type for the network parameter, you could probably do something like <source network='eth0'> and have it Just Work. Another model would be to have libvirt see an SR-IOV adapter as a network pool whereas it handled all of the VF management. Considering how inflexible SR-IOV is today, I'm not sure whether this is the best model. Has anyone put any more thought into this problem or how this should be modeled in libvirt? Michael, could you share your current thinking for -net syntax? -- Regards, Anthony Liguori

8 26

[libvirt] [PATCH] Let virt-manager 1.4.0+ work to access console of VM
by Bryan Quigley 19 Dec '17

19 Dec '17

Also see Ubuntu bug https://bugs.launchpad.net/ubuntu/+source/libvirt/+bug/1668681 First reported https://askubuntu.com/questions/833964/virt-manager-cant-connect-to-graphic… --- examples/apparmor/libvirt-qemu | 3 +++ 1 file changed, 3 insertions(+) diff --git a/examples/apparmor/libvirt-qemu b/examples/apparmor/libvirt-qemu index a9020aa..19d99e5 100644 --- a/examples/apparmor/libvirt-qemu +++ b/examples/apparmor/libvirt-qemu @@ -151,3 +151,6 @@ /etc/udev/udev.conf r, /sys/bus/ r, /sys/class/ r, + + # allow connect with openGraphicsFD to work + unix (send, receive) type=stream peer=(label=/usr/sbin/libvirtd), -- 2.10.2

3 5

[libvirt] [PATCH v2] qemu: command: align disk serial check to schema
by Nikolay Shirokovskiy 23 Aug '17

23 Aug '17

Disk serial schema has extra '.+' allowed characters in comparison with check in code. Looks like there is no reason for that as qemu allows any character AFAIK for serial. This discrepancy is originated in 85d15b51 where ability to add serial was added. --- Diff from v1: * fix xml2argv disk serial test to use all valid chars Looks like there is no existing infrastructure to test every invalid character. src/qemu/qemu_command.c | 2 +- tests/qemuxml2argvdata/qemuxml2argv-disk-serial.args | 2 +- tests/qemuxml2argvdata/qemuxml2argv-disk-serial.xml | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index c76f923..c5369b0 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -432,7 +432,7 @@ qemuBuildIoEventFdStr(virBufferPtr buf, } #define QEMU_SERIAL_PARAM_ACCEPTED_CHARS \ - "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789-_ " + "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789-_ .+" static int qemuSafeSerialParamValue(const char *value) diff --git a/tests/qemuxml2argvdata/qemuxml2argv-disk-serial.args b/tests/qemuxml2argvdata/qemuxml2argv-disk-serial.args index 2cefdca..fa0fc93 100644 --- a/tests/qemuxml2argvdata/qemuxml2argv-disk-serial.args +++ b/tests/qemuxml2argvdata/qemuxml2argv-disk-serial.args @@ -18,6 +18,6 @@ QEMU_AUDIO_DRV=none \ -boot c \ -usb \ -drive 'file=/dev/HostVG/QEMUGuest1,format=raw,if=none,id=drive-ide0-0-1,\ -serial=\ \ WD-WMAP9A966149' \ +serial=abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789-_\ .+' \ -device ide-drive,bus=ide.0,unit=1,drive=drive-ide0-0-1,id=ide0-0-1 \ -device virtio-balloon-pci,id=balloon0,bus=pci.0,addr=0x3 diff --git a/tests/qemuxml2argvdata/qemuxml2argv-disk-serial.xml b/tests/qemuxml2argvdata/qemuxml2argv-disk-serial.xml index 565462e..d54d73b 100644 --- a/tests/qemuxml2argvdata/qemuxml2argv-disk-serial.xml +++ b/tests/qemuxml2argvdata/qemuxml2argv-disk-serial.xml @@ -17,7 +17,7 @@ <disk type='block' device='disk'> <source dev='/dev/HostVG/QEMUGuest1'/> <target dev='hda' bus='ide'/> - <serial> WD-WMAP9A966149</serial> + <serial>abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789-_ .+</serial> <address type='drive' controller='0' bus='0' target='0' unit='1'/> </disk> <controller type='usb' index='0'/> -- 1.8.3.1

2 3

[libvirt] [PATCH RESEND] Interface: return appropriate status for bridge device
by Lin Ma 03 Aug '17

03 Aug '17

In function udevInterfaceIsActive, The current design relies on the sys attribute 'operstate' to determine whether the interface is active. For the device node of virtual network(say virbr0), There is already a dummy tap device to maintain a fixed mac on it, So it's available and its status should be considered as active. But if no anyelse tap device connects to it, The value of operstate of virbr0 in sysfs is 'down', So udevInterfaceIsActive returns 0, It causes 'virsh iface-list --all' to report the status of virbr0 as 'inactive'. This patch fixes the issue by counting slave devices for bridge device. Signed-off-by: Lin Ma <lma(a)suse.com> --- src/interface/interface_backend_udev.c | 22 ++++++++++++++++++++++ 1 file changed, 22 insertions(+) diff --git a/src/interface/interface_backend_udev.c b/src/interface/interface_backend_udev.c index 5d0fc64..9ac4674 100644 --- a/src/interface/interface_backend_udev.c +++ b/src/interface/interface_backend_udev.c @@ -1127,6 +1127,11 @@ udevInterfaceIsActive(virInterfacePtr ifinfo) struct udev_device *dev; virInterfaceDefPtr def = NULL; int status = -1; + struct dirent **member_list = NULL; + const char *devtype; + int member_count = 0; + char *member_path; + size_t i; dev = udev_device_new_from_subsystem_sysname(udev, "net", ifinfo->name); @@ -1146,6 +1151,23 @@ udevInterfaceIsActive(virInterfacePtr ifinfo) /* Check if it's active or not */ status = STREQ(udev_device_get_sysattr_value(dev, "operstate"), "up"); + if (!status) { + devtype = udev_device_get_devtype(dev); + if (STREQ_NULLABLE(devtype, "bridge")) { + if (virAsprintf(&member_path, "%s/%s", + udev_device_get_syspath(dev), "brif") < 0) + goto cleanup; + member_count = scandir(member_path, &member_list, + udevBridgeScanDirFilter, alphasort); + if (member_count > 0) + status = 1; + for (i = 0; i < member_count; i++) + VIR_FREE(member_list[i]); + VIR_FREE(member_list); + VIR_FREE(member_path); + } + } + udev_device_unref(dev); cleanup: -- 2.9.2

3 2

[libvirt] [PATCH v3 0/2] Allow saving VM state to pipe
by Chen Hanxiao 03 Aug '17

03 Aug '17

This series introduce flag VIR_DOMAIN_SAVE_DIRECT to enable command 'save' to write to PIPE. This will write QEMU_SAVE_MAGIC directly. Base upon patches from Roy Keene <rkeene(a)knightpoint.com> with some fixes. Change from original patch: 1) Check whether the specified path is a PIPE. 2) Rebase on upstream. 3) Add doc for virsh command v3: add doc/news.xml rebase on upstream v2-resend: rebase on upstream v2: rename VIR_DOMAIN_SAVE_PIPE to VIR_DOMAIN_SAVE_DIRECT remove S_ISFIFO check Chen Hanxiao (2): qemu: Allow qemuDomainSaveMemory saving VM state to a pipe virsh: introduce flage --direct for save command docs/news.xml | 9 +++++++ include/libvirt/libvirt-domain.h | 1 + src/qemu/qemu_driver.c | 54 ++++++++++++++++++++++++++-------------- tools/virsh-domain.c | 6 +++++ tools/virsh.pod | 5 +++- 5 files changed, 56 insertions(+), 19 deletions(-) -- 2.7.4

2 6

[libvirt] [PATCH 0/3] libxl: Add a test suite for libxl_domain_config generator
by Jim Fehlig 31 Jul '17

31 Jul '17

Long ago danpb posted some patches to test libvirt domXML to libxl_domain_config conversion https://www.redhat.com/archives/libvir-list/2014-May/msg01102.html Some of the prerequisite patches were pushed, but we've never managed to push patches actually providing the conversion tests. I sent several follow-ups to Dan's work but never converged on a satisfactory solution for all the Xen versions supported by libvirt. The last attempt was in Sept 2014 https://www.redhat.com/archives/libvir-list/2014-September/msg00698.html I tried to revive the work in Jan 2015, but that also stalled https://www.redhat.com/archives/libvir-list/2015-January/msg00924.html Fast-forward over 2.5 years from the first attempt and libvirt no longer supports older Xen versions 4.2 and 4.3 that were proving to be problematic. Starting with Xen 4.5 libxl added support for libxl_domain_config_from_json, which provides a way to implement the conversion tests that work with all Xen versions >= 4.5 (including latest xen.git master). The previous approaches compared a static json doc with the results of domXML->libxl_domain_config conversion done in the libxl driver. As discussed in those approaches, the json doc returned from libxl_domain_config_to_json can change over Xen releases as the libxl_domain_config object gains new fields, making it difficult to compare the conversion with a static doc. I had some time last week to pick at this thorn and reworked the test to make use of libxl_domain_config_from_json introduced in Xen 4.5. Instead of comparing the conversion results directly to a static json doc, the static doc is first round-tripped through _from_json -> _to_json. Any new fields added to libxl_domain_config object are then included in both docs, allowing comparison across multiple Xen releases. Patch3 provides the conversion tests using this new approach. The tests are not run on Xen 4.4 (oldest version currently supported by libvirt) since it does not provide libxl_domain_config_from_json. Patches 1 and 2 fix some issues found while working on the tests. See their commit messages for details. Jim Fehlig (3): libxl: determine device model version from emulator name libxl: relax checks on <emulator> libxl: Add a test suite for libxl_domain_config generator m4/virt-driver-libxl.m4 | 6 +- src/libxl/libxl_capabilities.c | 37 ++--- src/libxl/libxl_conf.c | 14 -- tests/Makefile.am | 18 ++- tests/libxlxml2domconfigdata/basic-hvm.json | 89 +++++++++++ tests/libxlxml2domconfigdata/basic-hvm.xml | 36 +++++ tests/libxlxml2domconfigdata/basic-pv.json | 65 ++++++++ tests/libxlxml2domconfigdata/basic-pv.xml | 28 ++++ tests/libxlxml2domconfigdata/moredevs-hvm.json | 111 +++++++++++++ tests/libxlxml2domconfigdata/moredevs-hvm.xml | 63 ++++++++ tests/libxlxml2domconfigtest.c | 208 +++++++++++++++++++++++++ tests/virmocklibxl.c | 87 +++++++++++ 12 files changed, 723 insertions(+), 39 deletions(-) create mode 100644 tests/libxlxml2domconfigdata/basic-hvm.json create mode 100644 tests/libxlxml2domconfigdata/basic-hvm.xml create mode 100644 tests/libxlxml2domconfigdata/basic-pv.json create mode 100644 tests/libxlxml2domconfigdata/basic-pv.xml create mode 100644 tests/libxlxml2domconfigdata/moredevs-hvm.json create mode 100644 tests/libxlxml2domconfigdata/moredevs-hvm.xml create mode 100644 tests/libxlxml2domconfigtest.c create mode 100644 tests/virmocklibxl.c -- 2.11.0

2 13

[libvirt] [PATCH v3 00/16] qemu: migration: show disks stats for nbd migration
by Nikolay Shirokovskiy 20 Jul '17

20 Jul '17

diff from v2: ============ 1. Fix style issues. 2. Rework patch for fetching job info (save logic to use temporary variable when drop vm lock) 3. Update disk stats when block jobs are canceled. 4. Adress a few more corner cases. This patch series add disks stats to domain job info(stats) as well as to migration completed event in case nbd scheme is used. There is little nuisance with qcow2 disks (which is the main scenario I guess) tied to the way qemu reports stats for this type of disks. For example if we have 64G disk filled only to 1G then stats start from 63G and will grow up to 64G on completion. The same way disk stats will be reported by this patch. I guess the better way to express the situation is to say we have 64G 'total', and have 'processed' field grow from 0G to 1G, like in case of memory stats. [1] is the example of completed memory stats of empty guest domain, which show difference between processed and total. There can be a workaround like getting initial blockjob offset position as a zero but is is rather ugly and racy and like uses undocumented behaviour. [1] memory migration stats example Memory processed: 3.307 MiB Memory remaining: 0.000 B Memory total: 1.032 GiB The above is applied to qemu 2.6 at least. Patches that were explicitly ACKed in previous review (up to style issues) marked with A. Nikolay Shirokovskiy (16): qemu: drop code for VIR_DOMAIN_JOB_BOUNDED and timeRemaining A qemu: introduce qemu domain job status A qemu: introduce QEMU_DOMAIN_JOB_STATUS_POSTCOPY A qemu: drop QEMU_MIGRATION_COMPLETED_UPDATE_STATS A qemu: drop excessive zero-out in qemuMigrationFetchJobStatus qemu: refactor fetching migration stats A qemu: simplify getting completed job stats qemu: fail querying destination migration statistics always qemu: start all async job with job status active qemu: introduce migrating job status qemu: always get job condition on getting job stats qemu: migrate: show disks stats on job info requests qemu: support getting disks stats during stopping block jobs qemu: migation: resolve race on getting job info and stopping block jobs qemu: migrate: copy disks stats to completed job qemu: migration: don't expose incomplete job as complete src/qemu/qemu_blockjob.c | 1 + src/qemu/qemu_domain.c | 38 +++++-- src/qemu/qemu_domain.h | 16 ++- src/qemu/qemu_driver.c | 95 ++++++++-------- src/qemu/qemu_migration.c | 236 ++++++++++++++++++++++++++------------- src/qemu/qemu_migration.h | 15 ++- src/qemu/qemu_migration_cookie.c | 7 +- src/qemu/qemu_monitor.c | 5 +- src/qemu/qemu_monitor.h | 4 +- src/qemu/qemu_monitor_json.c | 4 +- src/qemu/qemu_process.c | 10 +- tests/qemumonitorjsontest.c | 1 + 12 files changed, 273 insertions(+), 159 deletions(-) -- 1.8.3.1

2 19