- Devel - libvirt List Archives

[PATCH 0/3] qemu: migration: implement zstd compression
by Oleg Vasilev 20 Apr '23

20 Apr '23

QEMU now supports multifd-compression=zstd for migration with enabled multifd. Bring the support to libvirt as well. Claudio Fontana (1): qemu: migration: expose qemuMigrationParamsSetString Oleg Vasilev (2): qemu: migration: implement zstd compression tests: qemumigparams: test for zstd compression include/libvirt/libvirt-domain.h | 10 ++++ src/qemu/qemu_migration.h | 1 + src/qemu/qemu_migration_params.c | 79 +++++++++++++++++--------- src/qemu/qemu_migration_params.h | 7 +++ tests/qemumigparamsdata/zstd.json | 4 ++ tests/qemumigparamsdata/zstd.reply | 7 +++ tests/qemumigparamsdata/zstd.xml | 6 ++ tests/qemumigparamstest.c | 90 +++++++++++++++++++++++++++--- tools/virsh-completer-domain.c | 2 +- tools/virsh-domain.c | 14 +++++ 10 files changed, 187 insertions(+), 33 deletions(-) create mode 100644 tests/qemumigparamsdata/zstd.json create mode 100644 tests/qemumigparamsdata/zstd.reply create mode 100644 tests/qemumigparamsdata/zstd.xml -- 2.40.0

2 4

[libvirt PATCH 0/4] Various cleanups
by Ján Tomko 20 Apr '23

20 Apr '23

Found by cppcheck. Ján Tomko (4): conf: domain: remove unreachable break qemu: command: join two adjacent conditions ch: pinVcpuLive: remove unused variable conf: storage: remove redundant condition src/ch/ch_driver.c | 5 ----- src/conf/domain_conf.c | 1 - src/conf/storage_conf.c | 22 +++++++++------------- src/qemu/qemu_command.c | 2 -- 4 files changed, 9 insertions(+), 21 deletions(-) -- 2.39.2

2 5

[PATCH 0/2] Small fixes for restrictive numatune mode
by Martin Kletzander 20 Apr '23

20 Apr '23

See individual commit messages for explanations (duh!) Martin Kletzander (2): qemu, ch: Move threads to cgroup dir before changing parameters docs: Clarify restrictive numatune mode docs/formatdomain.rst | 11 ++++++++++- src/ch/ch_process.c | 12 ++++++------ src/qemu/qemu_process.c | 10 +++++----- 3 files changed, 21 insertions(+), 12 deletions(-) -- 2.40.0

3 4

[PATCH 0/3] docs improvements
by Peter Krempa 20 Apr '23

20 Apr '23

Peter Krempa (3): kbase: debuglogs: Emphasize disabling daemon timeout in 'TL;DR' section docs: manpages: Clarify that only TLS/TCP remote access needs 'virtproxyd' docs: manpages: State that TCP connection is insecure in 'virtproxyd' man page docs/kbase/debuglogs.rst | 6 ++++-- docs/manpages/virtbhyved.rst | 4 ++-- docs/manpages/virtinterfaced.rst | 4 ++-- docs/manpages/virtlxcd.rst | 4 ++-- docs/manpages/virtnetworkd.rst | 4 ++-- docs/manpages/virtnodedevd.rst | 4 ++-- docs/manpages/virtnwfilterd.rst | 4 ++-- docs/manpages/virtproxyd.rst | 3 +++ docs/manpages/virtqemud.rst | 4 ++-- docs/manpages/virtsecretd.rst | 4 ++-- docs/manpages/virtstoraged.rst | 4 ++-- docs/manpages/virtvboxd.rst | 4 ++-- docs/manpages/virtvzd.rst | 4 ++-- docs/manpages/virtxend.rst | 4 ++-- 14 files changed, 31 insertions(+), 26 deletions(-) -- 2.39.2

2 4

[PATCH 0/7] Various watchdog fixes
by Martin Kletzander 20 Apr '23

20 Apr '23

Los blurbos aqui. Martin Kletzander (7): conf: Add missing empty lines before virDomainWatchdogDefParseXML qemu: Fix grammar and quoting in watchdog error message on hotplug qemu: Forbid device attach of existing platform watchdog qemu: Forbid ib700 watchdogs for non-i440fx machine types qemu: Check all watchdogs for iTCO duplicates qemu: Validate watchdog action compatibility per-device qemu: Forbid most duplicated watchdogs src/conf/domain_conf.c | 2 + src/qemu/qemu_driver.c | 13 +++++ src/qemu/qemu_hotplug.c | 2 +- src/qemu/qemu_validate.c | 51 +++++++++++-------- tests/qemuhotplugtest.c | 3 ++ .../qemuhotplug-watchdog-reset.xml | 1 + ...plug-base-live+watchdog+watchdog-reset.xml | 1 + .../watchdog-q35-multiple.x86_64-latest.args | 4 +- .../watchdog-q35-multiple.xml | 2 +- .../watchdog-q35-multiple.x86_64-latest.xml | 4 +- 10 files changed, 58 insertions(+), 25 deletions(-) create mode 100644 tests/qemuhotplugtestdevices/qemuhotplug-watchdog-reset.xml create mode 120000 tests/qemuhotplugtestdomains/qemuhotplug-base-live+watchdog+watchdog-reset.xml -- 2.40.0

3 12

[libvirt PATCH] qemu: Allow sockets in long or deep paths.
by Nick Guenther 19 Apr '23

19 Apr '23

The qemu driver creates IPC sockets using absolute paths, but under POSIX socket paths are constrained pretty tightly. On systems with homedirs on an unusual mount point, like network homedirs, or just particularly long usernames, this could make starting VMs under qemu:///session impossible. Resolves https://gitlab.com/libvirt/libvirt/-/issues/466 Signed-off-by: Nick Guenther <nick.guenther(a)polymtl.ca> --- src/qemu/qemu_command.c | 52 ++++++++++++++++++++++++++++++++++++++--- 1 file changed, 49 insertions(+), 3 deletions(-) diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index 4ca93bf3dc..3f180d5fb6 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -53,6 +53,7 @@ #include "virmdev.h" #include "virutil.h" +#include <libgen.h> #include <sys/stat.h> #include <fcntl.h> @@ -4866,6 +4867,37 @@ qemuOpenChrChardevUNIXSocket(const virDomainChrSourceDef *dev) struct sockaddr_un addr; socklen_t addrlen = sizeof(addr); int fd; + char* wd = NULL; + char* socket_dir_c = NULL; + char* socket_name_c = NULL; + char* socket_dir = NULL; + char* socket_name = NULL; + + /* The path length is limited to what fits in sockaddr_un. + * It's pretty short: 108 on Linux, and this is too easy to hit. + * Work around this limit by using a *relative path*. + * + * background: https://stackoverflow.com/questions/34829600/why-is-the-maximal-path-length… + * + * docker added a different workaround: https://github.com/moby/moby/pull/13408 + */ + if ((wd = getcwd(NULL, 0)) == NULL) { + virReportSystemError(errno, "%s", + _("Unable to get working directory")); + goto error; + } + + socket_dir_c = strdup(dev->data.nix.path); // dirname edits the string given it, so it must be copied + socket_name_c = strdup(dev->data.nix.path); + + socket_dir = dirname(socket_dir_c); + socket_name = basename(socket_name_c); + + if (chdir(socket_dir) < 0) { + virReportSystemError(errno, "%s", + _("Unable to get change to socket directory")); + goto error; + } if ((fd = socket(AF_UNIX, SOCK_STREAM, 0)) < 0) { virReportSystemError(errno, "%s", @@ -4875,10 +4907,10 @@ qemuOpenChrChardevUNIXSocket(const virDomainChrSourceDef *dev) memset(&addr, 0, sizeof(addr)); addr.sun_family = AF_UNIX; - if (virStrcpyStatic(addr.sun_path, dev->data.nix.path) < 0) { + if (virStrcpyStatic(addr.sun_path, socket_name) < 0) { virReportError(VIR_ERR_INTERNAL_ERROR, - _("UNIX socket path '%1$s' too long"), - dev->data.nix.path); + _("UNIX socket name '%1$s' too long"), + socket_name); goto error; } @@ -4909,9 +4941,23 @@ qemuOpenChrChardevUNIXSocket(const virDomainChrSourceDef *dev) if (virFileUpdatePerm(dev->data.nix.path, 0002, 0664) < 0) goto error; + /* restore working directory */ + if (chdir(wd) < 0) { + virReportSystemError(errno, "%s", + _("Unable to restore working directory")); + goto error; + } + + free(socket_name_c); + free(socket_dir_c); + free(wd); + return fd; error: + if (socket_name_c != NULL) { free(socket_name_c); } + if (socket_dir_c != NULL) { free(socket_dir_c); } + if (wd != NULL) { free(wd); } VIR_FORCE_CLOSE(fd); return -1; } -- 2.34.1

5 6

[PATCH 0/3] qemu: hotplug: Fix media change in a hotplugged cdrom
by Peter Krempa 19 Apr '23

19 Apr '23

We need to update private data to handle it properly. Peter Krempa (3): qemuProcessRefreshDisks: Properly compare tray status qemuProcessRefreshDisks: Extract update of a single disk qemu: hotplug: Update disk private data after hotplug src/qemu/qemu_hotplug.c | 21 ++++++++++++---- src/qemu/qemu_process.c | 55 ++++++++++++++++++++++++----------------- src/qemu/qemu_process.h | 3 +++ tests/qemuhotplugtest.c | 40 ++++++++++++++++++++---------- 4 files changed, 79 insertions(+), 40 deletions(-) -- 2.39.2

2 4

[PATCH 0/2] qemu_domain: Increase memlock limit for NVMe disks
by Michal Privoznik 19 Apr '23

19 Apr '23

*** BLURB HERE *** Michal Prívozník (2): qemu_domain: Increase memlock limit for NVMe disks qemumemlocktest: Introduce pc-hostdev-nvme test case src/qemu/qemu_domain.c | 35 +++++++++++++------ .../qemumemlock-pc-hostdev-nvme.xml | 24 +++++++++++++ tests/qemumemlocktest.c | 1 + 3 files changed, 50 insertions(+), 10 deletions(-) create mode 100644 tests/qemumemlockdata/qemumemlock-pc-hostdev-nvme.xml -- 2.39.2

2 4

[PATCH v3 00/25] Setup iothread polling attributes in the XML
by Peter Krempa 19 Apr '23

19 Apr '23

Previous version: https://listman.redhat.com/archives/libvir-list/2023-March/239164.html Changes: - patches 1 - 23 are new and do the following - refactors to typed param handling 1 - 15, 18 - addition of new typed param APIs needed for changing the type we use to store the poll parameters currently, 16, 17, 19 - conversion of the existing parameter handling for iothread poll attributes to unsigned long long 20 - 23 - patches 24 and 25 are patches 1 and 2 of the previous posting: - use unsigned long long to store the values and adjust the code Peter Krempa (25): virTypedParameterAssignValue: Drop 'copystr' parameter util: virtypedparam: Use proper enum type for all switch() statements virTypedParamsDeserialize: Remove unnecessary line breaks virTypedParameterAssignValueVArgs: Ensure proper typed param type in caller util: virtypedparam: Simplify error handling in virTypedParamListAdd* virtypedparam.h: Consistently use contemporary header style util: virtypedparam: Introduce virTypedParamListNew() util: typedparam: Introduce 'virTypedParamListConcat' qemuDomainGetStatsBlock: Don't directly access virTypedParamList util: virtypedparam: Introduce 'virTypedParamListFetch' Use 'virTypedParamListFetch' for extracting identity parameters list util: virtypedparam: Privatize definition of struct _virTypedParamList util: virtypedparam: Refactor return value of virTypedParamListStealParams util: virtypedparam: Store errors inside virTypedParamList util: virtypedparam: Remove return values from virTypedParamListAdd* APIs util: typedparam: Introduce virTypedParamListAddUnsigned util: virtypedparam: Introduce virTypedParamsGetUnsigned virTypedParamsValidate: Refactor variable declaration and cleanup virTypedParamsValidate: Allow typed params to be both _UINT and _ULLONG virsh: cmdIOThreadSet: Refactor to use virTypedParamList qemu: Remove iothread 'poll-' value validation qemu: Store all iothread's 'poll*' attributes as unsigned long long virsh: cmdIOThreadSet: Use bigger types for --poll-grow and --poll-shrink conf: Store the iothread 'poll' settings in the XML qemu: Use configured iothread poll parameters on startup docs/formatdomain.rst | 11 +- include/libvirt/libvirt-domain.h | 4 +- src/admin/admin_server.c | 118 ++--- src/conf/domain_conf.c | 41 +- src/conf/domain_conf.h | 7 + src/conf/schemas/domaincommon.rng | 19 + src/driver.c | 7 +- src/libvirt-domain.c | 14 +- src/libvirt_private.syms | 7 +- src/qemu/qemu_command.c | 18 + src/qemu/qemu_domainjob.c | 49 +-- src/qemu/qemu_driver.c | 409 ++++++------------ src/qemu/qemu_monitor.h | 4 +- src/qemu/qemu_monitor_json.c | 48 +- src/remote/remote_daemon_dispatch.c | 8 +- src/test/test_driver.c | 34 +- src/util/virtypedparam.c | 401 +++++++++++------ src/util/virtypedparam.h | 187 +++++--- ...othreads-ids-pool-sizes.x86_64-latest.args | 6 +- .../iothreads-ids-pool-sizes.xml | 12 +- tools/virsh-domain.c | 83 ++-- 21 files changed, 781 insertions(+), 706 deletions(-) -- 2.39.2

3 44

[PATCH 0/3] A couple of network related fixes
by Michal Privoznik 19 Apr '23

19 Apr '23

*** BLURB HERE *** Michal Prívozník (3): networkUpdateState: do not assume dnsmasq_caps conf: Initialize _virNetworkObj::dnsmasqPid to -1 in virNetworkObjNew() networkRefreshDhcpDaemon: Get dnsmasq's PID once src/conf/virnetworkobj.c | 1 + src/network/bridge_driver.c | 7 +++++-- 2 files changed, 6 insertions(+), 2 deletions(-) -- 2.39.2

3 5

Re: [libvirt] [RFC] support vhost-user-scsi configuration
by wangjian 18 Apr '23

18 Apr '23

On 2023/4/18 21:50, wangjian (AN) wrote: > > On 4/15/23 15:48, wangjian (AN) wrote: >> Hi Guys, >> >> >> >> Currently qemu and spdk already support vhost-user-scsi, but there is >> no vhost-user-scsi configuration in libvirt. >> >> We hope that libvirt supports the following configurations to >> facilitate docking with qemu. >> >> >> >> <controller type='scsi' index='0' model='vhost-user-scsi'> >> >> <driver queues='4'/> >> >> <source type='unix' path='/var/tmp/scsi'> >> >> <reconnect enabled='yes' timeout='3'/> >> >> </source> >> >> <address type='pci' domain='0x0000' bus='0x02' slot='0x02' >> function='0x0'/> >> >> </controller> >> >> >> >> The usage in qemu like this: >> >> -chardev socket,id=chr-vu-virtio-disk10,path=/var/tmp/scsi,reconnect=3 >> >> -device >> vhost-user-blk-pci,num-queues=4,bus=pci.2,addr=0x0,chardev=chr-vu-virt >> io-disk10,id=scsi0 >> >> >> >> Could anyone give some suggestions? >> > > That's very likely because nobody wrote patches for it. I do not think there was a discussion where we decided to deliberately not support it. > But I guess, what advantage there is in letting an external helper manage a <controller/> ? I though, vhost-user-* is ideal for individual devices rather than controllers. > > Meanwhile, for testing purposes (definitely NOT production), you can use qemu:command line passthrough: > > https://libvirt.org/kbase/qemu-passthrough-security.html > > Michal > In spdk, a vhost-user-blk controller will only be associated with one disk, and a vhost-user-blk controller will only be associated with one socket file. Therefore, for the vhost-user-blk disk, it is appropriate to use the vhost-user-blk disk associated socket file configuration in libvirt. But in spdk, a vhost-user-scsi controller will be associated with multiple disks, and a vhost-user-scsi controller will only be associated with one socket file. Therefore, we think that for vhost-user-scsi, it is appropriate to use the controller to associate the socket file configuration in libvirt.

1 0

[libvirt PATCH] conf: Restrict use of <portForward> to the passt backend
by Andrea Bolognani 18 Apr '23

18 Apr '23

That's already the case in practice, but it's a better experience for the user if we reject this configuration outright instead of silently ignoring part of it. Signed-off-by: Andrea Bolognani <abologna(a)redhat.com> --- src/conf/domain_validate.c | 9 +++++++++ ...t-user-slirp-portforward.x86_64-latest.err | 1 + .../net-user-slirp-portforward.xml | 20 +++++++++++++++++++ tests/qemuxml2argvtest.c | 1 + 4 files changed, 31 insertions(+) create mode 100644 tests/qemuxml2argvdata/net-user-slirp-portforward.x86_64-latest.err create mode 100644 tests/qemuxml2argvdata/net-user-slirp-portforward.xml diff --git a/src/conf/domain_validate.c b/src/conf/domain_validate.c index ce6b8bf5a0..9c7ee6d75d 100644 --- a/src/conf/domain_validate.c +++ b/src/conf/domain_validate.c @@ -2097,6 +2097,15 @@ virDomainNetDefValidate(const virDomainNetDef *net) } } + if (net->nPortForwards > 0 && + (net->type != VIR_DOMAIN_NET_TYPE_USER || + (net->type == VIR_DOMAIN_NET_TYPE_USER && + net->backend.type != VIR_DOMAIN_NET_BACKEND_PASST))) { + virReportError(VIR_ERR_INTERNAL_ERROR, "%s", + _("The <portForward> element can only be used with <interface type='user'> and its 'passt' backend")); + return -1; + } + switch (net->type) { case VIR_DOMAIN_NET_TYPE_VHOSTUSER: if (!virDomainNetIsVirtioModel(net)) { diff --git a/tests/qemuxml2argvdata/net-user-slirp-portforward.x86_64-latest.err b/tests/qemuxml2argvdata/net-user-slirp-portforward.x86_64-latest.err new file mode 100644 index 0000000000..f296db1e8c --- /dev/null +++ b/tests/qemuxml2argvdata/net-user-slirp-portforward.x86_64-latest.err @@ -0,0 +1 @@ +internal error: The <portForward> element can only be used with <interface type='user'> and its 'passt' backend diff --git a/tests/qemuxml2argvdata/net-user-slirp-portforward.xml b/tests/qemuxml2argvdata/net-user-slirp-portforward.xml new file mode 100644 index 0000000000..721f04c878 --- /dev/null +++ b/tests/qemuxml2argvdata/net-user-slirp-portforward.xml @@ -0,0 +1,20 @@ +<domain type='qemu'> + <name>QEMUGuest1</name> + <uuid>c7a5fdbd-edaf-9455-926a-d65c16db1809</uuid> + <memory unit='KiB'>219136</memory> + <vcpu placement='static'>1</vcpu> + <os> + <type arch='x86_64' machine='pc'>hvm</type> + <boot dev='hd'/> + </os> + <devices> + <emulator>/usr/bin/qemu-system-x86_64</emulator> + <interface type='user'> + <mac address='00:11:22:33:44:55'/> + <portForward proto='tcp'> + <range start='443' to='344'/> + </portForward> + <model type='virtio'/> + </interface> + </devices> +</domain> diff --git a/tests/qemuxml2argvtest.c b/tests/qemuxml2argvtest.c index 1808d9fc02..23e0c4054c 100644 --- a/tests/qemuxml2argvtest.c +++ b/tests/qemuxml2argvtest.c @@ -1413,6 +1413,7 @@ mymain(void) DO_TEST_NOCAPS("net-user-addr"); DO_TEST_CAPS_LATEST("net-user-passt"); DO_TEST_CAPS_VER("net-user-passt", "7.2.0"); + DO_TEST_CAPS_LATEST_PARSE_ERROR("net-user-slirp-portforward"); DO_TEST_NOCAPS("net-virtio"); DO_TEST_NOCAPS("net-virtio-device"); DO_TEST_NOCAPS("net-virtio-disable-offloads"); -- 2.39.2

3 5

[PATCH 0/5] qemu: Fix chardev hotplug and deny <console/> live detach
by Michal Privoznik 18 Apr '23

18 Apr '23

While trying to make live detach of <console/> work, I've accumulated couple of patches. The most reasonable ones made it into this patch set as paches 1-4. And after nearly losing my sanity, I've decided it's not worth supporting live detach of <console/> since <serial/> works just fine. And that's what patch 5 does. If somebody thinks it's too harsh, I'm more than happy to review their patches that fix the problem. What problem? Well, for starters: https://bugzilla.redhat.com/show_bug.cgi?id=2156300#c4 https://bugzilla.redhat.com/show_bug.cgi?id=2156300#c8 Michal Prívozník (5): qemuDomainChrInsertPreAlloced: Fix adding implicit console qemuDomainChrRemove: Don't leak vmdef->consoles[0] qemuAssignDeviceChrAlias: Fix a crasher during <console/> hotplug qemuDomainRemoveChrDevice: Deal with qemuDomainChrRemove() failure qemu_hotplug: Deny live detach of <console/> src/qemu/qemu_alias.c | 1 + src/qemu/qemu_hotplug.c | 43 +++++++++++++++++++++++++++++------------ 2 files changed, 32 insertions(+), 12 deletions(-) -- 2.39.2

2 7

[PATCH] tests: viracpitest only works on little endian
by Boris Fiuczynski 18 Apr '23

18 Apr '23

Commit fc216db4fb789cbd309 introduced a mocked test with binary test data which fails on big endian machines. Therefore build the viracpitest test only on little endian machines. Fixes: fc216db4fb789cbd30917be036d0b94d965bdf7f Signed-off-by: Boris Fiuczynski <fiuczy(a)linux.ibm.com> --- tests/meson.build | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/tests/meson.build b/tests/meson.build index 35adbc2d56..0082446029 100644 --- a/tests/meson.build +++ b/tests/meson.build @@ -269,7 +269,6 @@ tests += [ { 'name': 'storagevolxml2xmltest' }, { 'name': 'sysinfotest' }, { 'name': 'utiltest' }, - { 'name': 'viracpitest' }, { 'name': 'viralloctest' }, { 'name': 'virauthconfigtest' }, { 'name': 'virbitmaptest' }, @@ -308,6 +307,12 @@ tests += [ { 'name': 'virmigtest' }, ] +if host_machine.endian() == 'little' + tests += [ + { 'name': 'viracpitest' }, + ] +endif + if host_machine.system() == 'linux' tests += [ { 'name': 'fchosttest' }, -- 2.39.0

3 2

[PATCH V2 0/3] Change default machine type for ARM and RISC-V
by Jim Fehlig 18 Apr '23

18 Apr '23

It is difficult, if not impossible, to create usable ARM and RISC-V VMs using the current default machine types of 'integratorcp' and 'spike_v1.10'. Change the default to the more suitable 'virt' type. See the following thread for more details https://listman.redhat.com/archives/libvir-list/2023-March/238746.html V1: https://listman.redhat.com/archives/libvir-list/2023-April/239392.html Changes in V2: Change the default for more ARM and RISC-V architectures Reword NEWS entry Jim Fehlig (3): qemu: Change default machine type for ARM qemu: Change default machine type for RISC-V NEWS: Mention change of default machine type for ARM and RISC-V NEWS.rst | 6 ++++++ src/qemu/qemu_capabilities.c | 10 +++++----- tests/domaincapsdata/qemu_4.2.0.aarch64.xml | 14 ++++++++++---- tests/domaincapsdata/qemu_5.0.0.aarch64.xml | 14 ++++++++++---- tests/domaincapsdata/qemu_5.2.0.aarch64.xml | 14 ++++++++++---- tests/domaincapsdata/qemu_6.0.0.aarch64.xml | 14 ++++++++++---- tests/domaincapsdata/qemu_6.2.0.aarch64.xml | 14 ++++++++++---- .../domaincapsdata/qemu_7.0.0-hvf.aarch64+hvf.xml | 6 ++++-- tests/domaincapsdata/qemu_7.0.0.aarch64.xml | 14 ++++++++++---- tests/qemucapabilitiesdata/caps_4.2.0_aarch64.xml | 4 ++-- tests/qemucapabilitiesdata/caps_5.0.0_aarch64.xml | 8 ++++---- tests/qemucapabilitiesdata/caps_5.0.0_riscv64.xml | 2 +- tests/qemucapabilitiesdata/caps_5.2.0_aarch64.xml | 8 ++++---- tests/qemucapabilitiesdata/caps_5.2.0_riscv64.xml | 2 +- tests/qemucapabilitiesdata/caps_6.0.0_aarch64.xml | 8 ++++---- tests/qemucapabilitiesdata/caps_6.2.0_aarch64.xml | 8 ++++---- .../caps_7.0.0_aarch64+hvf.xml | 8 ++++---- tests/qemucapabilitiesdata/caps_7.0.0_aarch64.xml | 8 ++++---- tests/qemucapabilitiesdata/caps_8.0.0_riscv64.xml | 4 ++-- 19 files changed, 105 insertions(+), 61 deletions(-) -- 2.40.0

2 5

[PATCH 0/3] Change default machine type for aarch64 and riscv64
by Jim Fehlig 18 Apr '23

18 Apr '23

This series is a first attempt at changing the default machine type for aarch64 and riscv64. It is difficult, if not impossible, to create a working VM with the current defaults of 'integratorcp' and 'spike_v1.10', so let's change the default to type 'virt'. See the following thread for more details https://listman.redhat.com/archives/libvir-list/2023-March/238746.html On aarch64, the default CPU for machine type 'virt' is 'cortex-a15', which works fine for <domain type='qemu'>, but fails for type='kvm' 2023-04-14T21:03:03.392256Z qemu-system-aarch64: KVM is not supported for this guest CPU type Do we need to go a step further and define a default CPU (presumably host-passthrough) for machine type 'virt' for kvm domains on aarch64? TIA for comments and suggestions! Regards, Jim Jim Fehlig (3): qemu: Change default machine type for aarch64 qemu: Change default machine type for riscv64 NEWS: Mention change of default machine type for aarch64 and riscv64 NEWS.rst | 6 ++++++ src/qemu/qemu_capabilities.c | 4 ++-- tests/domaincapsdata/qemu_4.2.0.aarch64.xml | 14 ++++++++++---- tests/domaincapsdata/qemu_5.0.0.aarch64.xml | 14 ++++++++++---- tests/domaincapsdata/qemu_5.2.0.aarch64.xml | 14 ++++++++++---- tests/domaincapsdata/qemu_6.0.0.aarch64.xml | 14 ++++++++++---- tests/domaincapsdata/qemu_6.2.0.aarch64.xml | 14 ++++++++++---- .../domaincapsdata/qemu_7.0.0-hvf.aarch64+hvf.xml | 6 ++++-- tests/domaincapsdata/qemu_7.0.0.aarch64.xml | 14 ++++++++++---- tests/qemucapabilitiesdata/caps_4.2.0_aarch64.xml | 4 ++-- tests/qemucapabilitiesdata/caps_5.0.0_aarch64.xml | 8 ++++---- tests/qemucapabilitiesdata/caps_5.0.0_riscv64.xml | 2 +- tests/qemucapabilitiesdata/caps_5.2.0_aarch64.xml | 8 ++++---- tests/qemucapabilitiesdata/caps_5.2.0_riscv64.xml | 2 +- tests/qemucapabilitiesdata/caps_6.0.0_aarch64.xml | 8 ++++---- tests/qemucapabilitiesdata/caps_6.2.0_aarch64.xml | 8 ++++---- .../caps_7.0.0_aarch64+hvf.xml | 8 ++++---- tests/qemucapabilitiesdata/caps_7.0.0_aarch64.xml | 8 ++++---- tests/qemucapabilitiesdata/caps_8.0.0_riscv64.xml | 4 ++-- 19 files changed, 102 insertions(+), 58 deletions(-) -- 2.40.0

2 8

[PATCH] kbase: debuglogs: Add a 'TL; DR' section for enabling logging in most common case
by Peter Krempa 17 Apr '23

17 Apr '23

The document grew a bit too much explaining all the mistakes we've seen the users do when configuring logging. Add a section distilling the configuration of the most basic scenario which we can refer to when upstream issues are reported. The scenario is for a runtime setting of logging into a file applied to the 'virtqemud' daemon. Signed-off-by: Peter Krempa <pkrempa(a)redhat.com> --- docs/kbase/debuglogs.rst | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) diff --git a/docs/kbase/debuglogs.rst b/docs/kbase/debuglogs.rst index 53d70ee748..811ccf0102 100644 --- a/docs/kbase/debuglogs.rst +++ b/docs/kbase/debuglogs.rst @@ -16,6 +16,22 @@ server side that matters as nearly all interesting work takes place there. Moreover, libvirt catches stderr of all running domains. These can be useful as well. +TL;DR - Enable debug logs for most common scenario +=================================================== + +This applies to the most common scenario of ``system`` instance of +``virtqemud``. Log setting is not persisted, so a restart of ``virtqemud`` or +the system clears this setting:: + + # virt-admin -c virtqemud:///system daemon-log-outputs "3:journald 1:file:/var/log/libvirt/libvirtd.log" + # virt-admin -c virtqemud:///system daemon-log-filters "3:remote 4:event 3:util.json 3:util.object 3:util.dbus 3:util.netlink 3:node_device 3:rpc 3:access 1:*" + + # # optionally disable timeout of the daemon + # virt-admin -c virtqemud:///system daemon-timeout 0 + +For any other configuration please read the rest of the document. If you want +to persist the log level and log outputs settings edit +``/etc/libvirt/virtqemud.conf`` and look for ``log-filters`` and ``log-outputs`` Logging settings in libvirt =========================== -- 2.39.2

3 3

[PATCH 0/2] virsh: Introduce --xpath and --wrap to (dom)capabilities
by Michal Privoznik 17 Apr '23

17 Apr '23

*** BLURB HERE *** Michal Prívozník (2): virsh: Introduce --xpath and --wrap to capabilities virsh: Introduce --xpath and --wrap to domcapabilities docs/manpages/virsh.rst | 17 ++++++++++++++- tools/virsh-host.c | 46 ++++++++++++++++++++++++++++++++++------- 2 files changed, 54 insertions(+), 9 deletions(-) -- 2.39.2

2 3

[PATCH 0/3] Setup iothread polling attributes in the XML
by Peter Krempa 17 Apr '23

17 Apr '23

Peter Krempa (3): conf: Store the iothread 'poll' settings in the XML qemu: Use configured iothread poll parameters on startup docs: formatdomain: Properly indent example XML for setting 'metadata_cache' docs/formatdomain.rst | 53 +++++++++++-------- src/conf/domain_conf.c | 41 +++++++++++++- src/conf/domain_conf.h | 7 +++ src/conf/schemas/domaincommon.rng | 19 +++++++ src/qemu/qemu_command.c | 18 +++++++ src/qemu/qemu_driver.c | 30 ++++++----- ...othreads-ids-pool-sizes.x86_64-latest.args | 6 +-- .../iothreads-ids-pool-sizes.xml | 12 +++-- 8 files changed, 142 insertions(+), 44 deletions(-) -- 2.39.2

2 9

Plans for the next release
by Jiri Denemark 17 Apr '23

17 Apr '23

We are getting close to the next release of libvirt. To aim for the release on May 02 I suggest entering the freeze on Tuesday Apr 25 and tagging RC2 on Thursday Apr 27. I hope this works for everyone. Jirka

1 0

[libvirt PATCH] network: do not assume dnsmasq in networkUpdateState
by Ján Tomko 17 Apr '23

17 Apr '23

If we don't have dnsmasq, it's pointless to try to find its pidfile. Also, dnsmasqCapsGetBinaryPath would access a NULL pointer. Fixes: 4b68c982e283471575bacbf87302495864da46fe Foxes: https://gitlab.com/libvirt/libvirt/-/issues/456 Signed-off-by: Ján Tomko <jtomko(a)redhat.com> --- src/network/bridge_driver.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/network/bridge_driver.c b/src/network/bridge_driver.c index 3fa56bfc09..ee4bbd4a93 100644 --- a/src/network/bridge_driver.c +++ b/src/network/bridge_driver.c @@ -492,7 +492,7 @@ networkUpdateState(virNetworkObj *obj, virNetworkObjPortForEach(obj, networkUpdatePort, obj); /* Try and read dnsmasq pids of active networks */ - if (virNetworkObjIsActive(obj) && def->ips && (def->nips > 0)) { + if (dnsmasq_caps && virNetworkObjIsActive(obj) && def->ips && (def->nips > 0)) { pid_t dnsmasqPid; if (networkSetMacMap(cfg, obj) < 0) -- 2.39.2

3 5

[libvirt] [RFC] support vhost-user-scsi configuration
by wangjian (AN) 17 Apr '23

17 Apr '23

Hi Guys, Currently qemu and spdk already support vhost-user-scsi, but there is no vhost-user-scsi configuration in libvirt. We hope that libvirt supports the following configurations to facilitate docking with qemu. <controller type='scsi' index='0' model='vhost-user-scsi'> <driver queues='4'/> <source type='unix' path='/var/tmp/scsi'> <reconnect enabled='yes' timeout='3'/> </source> <address type='pci' domain='0x0000' bus='0x02' slot='0x02' function='0x0'/> </controller> The usage in qemu like this: -chardev socket,id=chr-vu-virtio-disk10,path=/var/tmp/scsi,reconnect=3 -device vhost-user-blk-pci,num-queues=4,bus=pci.2,addr=0x0,chardev=chr-vu-virtio-disk10,id=scsi0 Could anyone give some suggestions?

2 1

[PATCH v2] qemu: Fix domxml-to-native command failure
by Zhenzhong Duan 17 Apr '23

17 Apr '23

virsh command domxml-to-native failed with below error but start command succeed for same domain xml. "internal error: invalid PCI passthrough type 'default'" If host PCI device's driver attribute isn't defined in domain xml, qemu driver will choose a proper value based on host environment before starting qemu process. But this is missed in domxml-to-native command, add the same logic so domxml-to-native could pass and replace the relevant code in qemuHostdevPreparePCIDevicesCheckSupport() with an error report. Signed-off-by: Zhenzhong Duan <zhenzhong.duan(a)intel.com> --- v2: replace setting with an error report, suggested by Michal src/qemu/qemu_domain.c | 11 +++++++++++ src/qemu/qemu_hostdev.c | 8 ++++---- 2 files changed, 15 insertions(+), 4 deletions(-) diff --git a/src/qemu/qemu_domain.c b/src/qemu/qemu_domain.c index 63b13b6875c9..517c90f93f0a 100644 --- a/src/qemu/qemu_domain.c +++ b/src/qemu/qemu_domain.c @@ -11274,6 +11274,17 @@ qemuDomainPrepareHostdev(virDomainHostdevDef *hostdev, } } } + if (hostdev->mode == VIR_DOMAIN_HOSTDEV_MODE_SUBSYS && + hostdev->source.subsys.type == VIR_DOMAIN_HOSTDEV_SUBSYS_TYPE_PCI) { + bool supportsPassthroughVFIO = qemuHostdevHostSupportsPassthroughVFIO(); + virDomainHostdevSubsysPCIBackendType *backend = &hostdev->source.subsys.u.pci.backend; + + if (*backend == VIR_DOMAIN_HOSTDEV_PCI_BACKEND_DEFAULT && + supportsPassthroughVFIO && + virQEMUCapsGet(priv->qemuCaps, QEMU_CAPS_DEVICE_VFIO_PCI)) { + *backend = VIR_DOMAIN_HOSTDEV_PCI_BACKEND_VFIO; + } + } return 0; } diff --git a/src/qemu/qemu_hostdev.c b/src/qemu/qemu_hostdev.c index 45cd1066f0a5..305a323a0769 100644 --- a/src/qemu/qemu_hostdev.c +++ b/src/qemu/qemu_hostdev.c @@ -175,15 +175,15 @@ qemuHostdevPreparePCIDevicesCheckSupport(virDomainHostdevDef **hostdevs, case VIR_DOMAIN_HOSTDEV_PCI_BACKEND_DEFAULT: if (supportsPassthroughVFIO && virQEMUCapsGet(qemuCaps, QEMU_CAPS_DEVICE_VFIO_PCI)) { - *backend = VIR_DOMAIN_HOSTDEV_PCI_BACKEND_VFIO; + virReportError(VIR_ERR_INTERNAL_ERROR, "%s", + _("qemuDomainPrepareHostdev() should have set backend " + "to VIR_DOMAIN_HOSTDEV_PCI_BACKEND_VFIO")); } else { virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", _("host doesn't support passthrough of " "host PCI devices")); - return false; } - - break; + return false; case VIR_DOMAIN_HOSTDEV_PCI_BACKEND_VFIO: if (!supportsPassthroughVFIO) { -- 2.25.1

3 2

Improve default machine type selection
by Jim Fehlig 14 Apr '23

14 Apr '23

If an explicit machine type is not specified in the VM config, the qemu driver will select the first machine type in the list of machine types for the specified accelerator. See virQEMUCapsGetPreferredMachine https://gitlab.com/libvirt/libvirt/-/blob/master/src/qemu/qemu_capabilities… On my test machines, this works reasonably well for x86_64 where the first machine type is pc-i440fx-7.1. But for aarch64, the first machine is integratorcp, which is not very useful with maxCpus=1 and other limitations. Would it be possible to select a "better" default machine type? E.g. 'pc' for x86_64, 'virt' for aarch64, etc. I'm happy to work on this if folks deem it's a reasonable improvement. It would benefit tools like the libvirt terraform provider, which prefer to use something more generic like <type>hvm</type>. Regards, Jim

4 18

[PATCH] qemu: Fix incorrect command name in error messages
by Andrea Bolognani 14 Apr '23

14 Apr '23

Signed-off-by: Andrea Bolognani <abologna(a)redhat.com> --- Pushed as trivial. src/qemu/qemu_monitor_json.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/src/qemu/qemu_monitor_json.c b/src/qemu/qemu_monitor_json.c index 3b42b28fa1..b2c0b20a11 100644 --- a/src/qemu/qemu_monitor_json.c +++ b/src/qemu/qemu_monitor_json.c @@ -4813,7 +4813,7 @@ int qemuMonitorJSONGetMachines(qemuMonitor *mon, if (virJSONValueObjectHasKey(child, "numa-mem-supported")) { if (virJSONValueObjectGetBoolean(child, "numa-mem-supported", &info->numaMemSupported) < 0) { virReportError(VIR_ERR_INTERNAL_ERROR, "%s", - _("qemu-machines reply has malformed " + _("query-machines reply has malformed " "'numa-mem-supported' data")); goto cleanup; } @@ -4841,7 +4841,7 @@ int qemuMonitorJSONGetMachines(qemuMonitor *mon, if (virJSONValueObjectGetBoolean(child, "acpi", &acpi) < 0) { virReportError(VIR_ERR_INTERNAL_ERROR, "%s", - _("qemu-machines reply has malformed 'acpi data")); + _("query-machines reply has malformed 'acpi data")); goto cleanup; } -- 2.39.2

1 0