August 2020 - Devel - libvirt List Archives

[RFC] qemu: Check for changes in qemu modules directory
by Jim Fehlig 26 Aug '20

26 Aug '20

Add a configuration option for specifying location of the qemu modules directory, defaulting to /usr/lib64/qemu. Then use this location to check for changes in the directory, indicating that a qemu module has changed and capabilities need to be reprobed. Signed-off-by: Jim Fehlig <jfehlig(a)suse.com> --- This is a mildly tested impl of the idea discussed here (and elsewhere in the thread) https://www.redhat.com/archives/libvir-list/2020-August/msg00684.html I've checked that both system and session caches are updated when a qemu modules package is updated. I'll do some more testing but wanted to send the patch for early comments. Testing in other environments would be much appreciated. Thanks! meson.build | 6 ++++++ meson_options.txt | 1 + src/qemu/qemu_capabilities.c | 35 +++++++++++++++++++++++++++++++++++ 3 files changed, 42 insertions(+) diff --git a/meson.build b/meson.build index a72d0c0e85..16928482aa 100644 --- a/meson.build +++ b/meson.build @@ -1745,6 +1745,12 @@ if not get_option('driver_qemu').disabled() if use_qemu conf.set('WITH_QEMU', 1) + qemu_moddir = get_option('qemu_moddir') + if qemu_moddir == '' + qemu_moddir = '/usr/lib64/qemu' + endif + conf.set_quoted('QEMU_MODDIR', qemu_moddir) + if host_machine.system() in ['freebsd', 'darwin'] default_qemu_user = 'root' default_qemu_group = 'wheel' diff --git a/meson_options.txt b/meson_options.txt index c538d323c1..cedbd79491 100644 --- a/meson_options.txt +++ b/meson_options.txt @@ -59,6 +59,7 @@ option('driver_openvz', type: 'feature', value: 'auto', description: 'OpenVZ dri option('driver_qemu', type: 'feature', value: 'auto', description: 'QEMU/KVM driver') option('qemu_user', type: 'string', value: '', description: 'username to run QEMU system instance as') option('qemu_group', type: 'string', value: '', description: 'groupname to run QEMU system instance as') +option('qemu_moddir', type: 'string', value: '', description: 'set the directory where QEMU modules are located') option('driver_remote', type: 'feature', value: 'enabled', description: 'remote driver') option('remote_default_mode', type: 'combo', choices: ['legacy', 'direct'], value: 'legacy', description: 'remote driver default mode') option('driver_secrets', type: 'feature', value: 'auto', description: 'local secrets management driver') diff --git a/src/qemu/qemu_capabilities.c b/src/qemu/qemu_capabilities.c index ff6ba8c9e9..a00853c753 100644 --- a/src/qemu/qemu_capabilities.c +++ b/src/qemu/qemu_capabilities.c @@ -677,6 +677,7 @@ struct _virQEMUCaps { char *binary; time_t ctime; time_t libvirtCtime; + time_t modDirMtime; bool invalidation; virBitmapPtr flags; @@ -4194,6 +4195,7 @@ virQEMUCapsParseSEVInfo(virQEMUCapsPtr qemuCaps, xmlXPathContextPtr ctxt) * <qemuCaps> * <emulator>/some/path</emulator> * <qemuctime>234235253</qemuctime> + * <qemumoddirmtime>234235253</qemumoddirmtime> * <selfctime>234235253</selfctime> * <selfvers>1002016</selfvers> * <flag name='foo'/> @@ -4283,6 +4285,9 @@ virQEMUCapsLoadCache(virArch hostArch, } qemuCaps->ctime = (time_t)l; + if (virXPathLongLong("string(./qemumoddirmtime)", ctxt, &l) == 0) + qemuCaps->modDirMtime = (time_t)l; + if ((n = virXPathNodeSet("./flag", ctxt, &nodes)) < 0) { virReportError(VIR_ERR_INTERNAL_ERROR, "%s", _("failed to parse qemu capabilities flags")); @@ -4615,6 +4620,10 @@ virQEMUCapsFormatCache(virQEMUCapsPtr qemuCaps) qemuCaps->binary); virBufferAsprintf(&buf, "<qemuctime>%llu</qemuctime>\n", (long long)qemuCaps->ctime); + if (qemuCaps->modDirMtime > 0) { + virBufferAsprintf(&buf, "<qemumoddirmtime>%llu</qemumoddirmtime>\n", + (long long)qemuCaps->modDirMtime); + } virBufferAsprintf(&buf, "<selfctime>%llu</selfctime>\n", (long long)qemuCaps->libvirtCtime); virBufferAsprintf(&buf, "<selfvers>%lu</selfvers>\n", @@ -4881,6 +4890,23 @@ virQEMUCapsIsValid(void *data, if (!qemuCaps->binary) return true; + if (virFileExists(QEMU_MODDIR)) { + if (stat(QEMU_MODDIR, &sb) < 0) { + VIR_DEBUG("Failed to stat QEMU module directory '%s': %s", + QEMU_MODDIR, + g_strerror(errno)); + return false; + } + + if (sb.st_mtime != qemuCaps->modDirMtime) { + VIR_DEBUG("Outdated capabilities for '%s': QEMU modules " + "directory '%s' changed (%lld vs %lld)", + qemuCaps->binary, QEMU_MODDIR, + (long long)sb.st_mtime, (long long)qemuCaps->modDirMtime); + return false; + } + } + if (qemuCaps->libvirtCtime != virGetSelfLastChanged() || qemuCaps->libvirtVersion != LIBVIR_VERSION_NUMBER) { VIR_DEBUG("Outdated capabilities for '%s': libvirt changed " @@ -5463,6 +5489,15 @@ virQEMUCapsNewForBinaryInternal(virArch hostArch, goto error; } + if (virFileExists(QEMU_MODDIR)) { + if (stat(QEMU_MODDIR, &sb) < 0) { + virReportSystemError(errno, _("Cannot check QEMU module directory %s"), + QEMU_MODDIR); + goto error; + } + qemuCaps->modDirMtime = sb.st_mtime; + } + if (virQEMUCapsInitQMP(qemuCaps, libDir, runUid, runGid) < 0) goto error; -- 2.28.0

2 4

[PATCH] apparmor: allow libvirtd to call virtiofsd
by Kevin Locke 26 Aug '20

26 Aug '20

When using [virtiofs], libvirtd must launch [virtiofsd] to provide filesystem access on the host. When a guest is configured with virtiofs, such as: <filesystem type='mount' accessmode='passthrough'> <driver type='virtiofs'/> <source dir='/path'/> <target dir='mount_tag'/> </filesystem> Attempting to start the guest fails with: internal error: virtiofsd died unexpectedly /var/log/libvirt/qemu/$name-fs0-virtiofsd.log contains: libvirt: error : cannot execute binary /usr/lib/qemu/virtiofsd: Permission denied dmesg contains: audit: type=1400 audit(1598229295.959:73): apparmor="DENIED" operation="exec" profile="libvirtd" name="/usr/lib/qemu/virtiofsd" pid=46007 comm="rpc-worker" requested_mask="x" denied_mask="x" fsuid=0 ouid=0 To avoid this, allow execution of virtiofsd from the libvirtd AppArmor profile. [virtiofs]: https://libvirt.org/kbase/virtiofs.html [virtiofsd]: https://www.qemu.org/docs/master/interop/virtiofsd.html Signed-off-by: Kevin Locke <kevin(a)kevinlocke.name> --- src/security/apparmor/usr.sbin.libvirtd.in | 1 + 1 file changed, 1 insertion(+) diff --git a/src/security/apparmor/usr.sbin.libvirtd.in b/src/security/apparmor/usr.sbin.libvirtd.in index 4518e8f865..f2030764cd 100644 --- a/src/security/apparmor/usr.sbin.libvirtd.in +++ b/src/security/apparmor/usr.sbin.libvirtd.in @@ -89,6 +89,7 @@ profile libvirtd @sbindir@/libvirtd flags=(attach_disconnected) { /usr/lib/xen-*/bin/libxl-save-helper PUx, /usr/lib/xen-*/bin/pygrub PUx, /usr/{lib,lib64,lib/qemu,libexec}/vhost-user-gpu PUx, + /usr/{lib,lib64,lib/qemu,libexec}/virtiofsd PUx, # Required by nwfilter_ebiptables_driver.c:ebiptablesWriteToTempFile() to # read and run an ebtables script. -- 2.28.0

3 6

[PATCH V4 0/3] Xen: Support PCI permissive setting with writeFiltering attribute
by Jim Fehlig 26 Aug '20

26 Aug '20

This is V4 of Marek's series to support the xl.cfg(5) permissive setting on PCI devices. Previous versions of the series V3 https://www.redhat.com/archives/libvir-list/2020-August/msg00465.html V2 https://www.redhat.com/archives/libvir-list/2020-April/msg01230.html Changes from V3: - Add a check to qemu_validate to report error if writeFiltering is used in a qemu domain. - Rebase to master. Changes from V2: - Instead of using a permisssive attribute on the <hostdev> element, use a writeFiltering attribute on the <hostdev>'s <source> element. Rational being that the filtering of writes to the PCI config space is done at the source. Jim Fehlig (3): Xen: Add writeFiltering option for PCI devices Xen: Add support for writeFiltering in config converter News: Advertise support for writeFiltering attribute of PCI hostdevs NEWS.rst | 7 ++ docs/formatdomain.rst | 7 +- docs/schemas/domaincommon.rng | 5 ++ src/conf/domain_conf.c | 14 ++++ src/conf/domain_conf.h | 1 + src/libxl/libxl_conf.c | 1 + src/libxl/xen_common.c | 77 +++++++++++++++---- src/qemu/qemu_validate.c | 7 ++ .../libxlxml2domconfigdata/moredevs-hvm.json | 6 ++ tests/libxlxml2domconfigdata/moredevs-hvm.xml | 5 ++ tests/xlconfigdata/test-fullvirt-pci.cfg | 25 ++++++ tests/xlconfigdata/test-fullvirt-pci.xml | 53 +++++++++++++ tests/xlconfigtest.c | 1 + 13 files changed, 195 insertions(+), 14 deletions(-) create mode 100644 tests/xlconfigdata/test-fullvirt-pci.cfg create mode 100644 tests/xlconfigdata/test-fullvirt-pci.xml -- 2.28.0

2 6

[libvirt PATCH 0/9] qemu: Allow migration over UNIX sockets
by Martin Kletzander 26 Aug '20

26 Aug '20

KubeVirt would like to use this feature. For more information see individual commits and changes in manpages and documentation. Resolves: https://bugzilla.redhat.com/1638889 Martin Kletzander (9): qemu: Use g_autofree in qemuMigrationSrcConnect qemu: Rework qemuMigrationSrcConnect virsh: Reuse existing variable when parsing migrate --disks-port qemu: Rework starting NBD server for migration tests: Add simple test for virDomainMigrateCheckNotLocal qemu: Allow NBD migration over UNIX socket peer2peer migration: allow connecting to local sockets qemu: Allow migration over UNIX socket news: qemu: Allow migration over UNIX sockets NEWS.rst | 6 + docs/manpages/virsh.rst | 30 +++- docs/migration.html.in | 33 ++++ include/libvirt/libvirt-domain.h | 12 ++ scripts/apibuild.py | 1 + src/libvirt-domain.c | 12 +- src/libvirt_internal.h | 2 + src/libvirt_private.syms | 1 + src/qemu/qemu_domain.h | 1 + src/qemu/qemu_driver.c | 33 +++- src/qemu/qemu_migration.c | 284 ++++++++++++++++++++++--------- src/qemu/qemu_migration.h | 3 + src/qemu/qemu_migration_cookie.c | 22 ++- src/qemu/qemu_migration_cookie.h | 1 + src/remote/remote_driver.c | 8 +- src/util/viruri.c | 30 ++++ src/util/viruri.h | 2 + tests/meson.build | 1 + tests/virmigtest.c | 90 ++++++++++ tools/virsh-domain.c | 19 ++- 20 files changed, 483 insertions(+), 108 deletions(-) create mode 100644 tests/virmigtest.c -- 2.28.0

3 29

[GSoC][PATCH v2] virmigraiton: `qemuMigrationJobPhase` transformed for more generic use
by Prathamesh Chavan 26 Aug '20

26 Aug '20

`qemuMigrationJobPhase` was transformed into `virMigrationJobPhase` and a common util file `virmigration` was created to store its defination. This is one of the initial steps we are taking towards making more and more code hypervisor agnostic. And this shall be followed by transformation of similar structs in the future. Signed-off-by: Prathamesh Chavan <pc44800(a)gmail.com> --- Previous version of this patch can be found here[1]. After recent addition of `qemu_snapshot` files, I thought about updating the patch on the new master so that checking it out will be easier. [1]: https://www.redhat.com/archives/libvir-list/2020-August/msg00743.html src/hypervisor/meson.build | 1 + src/hypervisor/virmigration.c | 41 +++++++++++++++++++ src/hypervisor/virmigration.h | 38 ++++++++++++++++++ src/libvirt_private.syms | 4 ++ src/qemu/MIGRATION.txt | 8 ++-- src/qemu/qemu_domainjob.c | 4 +- src/qemu/qemu_migration.c | 74 +++++++++++++++++------------------ src/qemu/qemu_migration.h | 17 +------- src/qemu/qemu_process.c | 48 +++++++++++------------ 9 files changed, 151 insertions(+), 84 deletions(-) create mode 100644 src/hypervisor/virmigration.c create mode 100644 src/hypervisor/virmigration.h diff --git a/src/hypervisor/meson.build b/src/hypervisor/meson.build index 85149c683e..c81bdfa2fc 100644 --- a/src/hypervisor/meson.build +++ b/src/hypervisor/meson.build @@ -3,6 +3,7 @@ hypervisor_sources = [ 'domain_driver.c', 'virclosecallbacks.c', 'virhostdev.c', + 'virmigration.c', ] hypervisor_lib = static_library( diff --git a/src/hypervisor/virmigration.c b/src/hypervisor/virmigration.c new file mode 100644 index 0000000000..2cad5a6b1b --- /dev/null +++ b/src/hypervisor/virmigration.c @@ -0,0 +1,41 @@ +/* + * virmigration.c: hypervisor migration handling + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library. If not, see + * <http://www.gnu.org/licenses/>. + */ + +#include <config.h> + +#include "virmigration.h" +#include "domain_driver.h" +#include "virlog.h" + +#define VIR_FROM_THIS VIR_FROM_DOMAIN + +VIR_LOG_INIT("util.migration"); + +VIR_ENUM_IMPL(virMigrationJobPhase, + VIR_MIGRATION_PHASE_LAST, + "none", + "perform2", + "begin3", + "perform3", + "perform3_done", + "confirm3_cancelled", + "confirm3", + "prepare", + "finish2", + "finish3", +); diff --git a/src/hypervisor/virmigration.h b/src/hypervisor/virmigration.h new file mode 100644 index 0000000000..e03d71c1bb --- /dev/null +++ b/src/hypervisor/virmigration.h @@ -0,0 +1,38 @@ +/* + * virmigration.h: hypervisor migration handling + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library. If not, see + * <http://www.gnu.org/licenses/>. + */ + +#pragma once + +#include "virenum.h" + + +typedef enum { + VIR_MIGRATION_PHASE_NONE = 0, + VIR_MIGRATION_PHASE_PERFORM2, + VIR_MIGRATION_PHASE_BEGIN3, + VIR_MIGRATION_PHASE_PERFORM3, + VIR_MIGRATION_PHASE_PERFORM3_DONE, + VIR_MIGRATION_PHASE_CONFIRM3_CANCELLED, + VIR_MIGRATION_PHASE_CONFIRM3, + VIR_MIGRATION_PHASE_PREPARE, + VIR_MIGRATION_PHASE_FINISH2, + VIR_MIGRATION_PHASE_FINISH3, + + VIR_MIGRATION_PHASE_LAST +} virMigrationJobPhase; +VIR_ENUM_DECL(virMigrationJobPhase); diff --git a/src/libvirt_private.syms b/src/libvirt_private.syms index f950a68179..c7adf16aba 100644 --- a/src/libvirt_private.syms +++ b/src/libvirt_private.syms @@ -1477,6 +1477,10 @@ virHostdevUpdateActiveSCSIDevices; virHostdevUpdateActiveUSBDevices; +# hypervisor/virmigration.h +virMigrationJobPhaseTypeFromString; +virMigrationJobPhaseTypeToString; + # libvirt_internal.h virConnectSupportsFeature; virDomainMigrateBegin3; diff --git a/src/qemu/MIGRATION.txt b/src/qemu/MIGRATION.txt index e861fd001e..dd044c6064 100644 --- a/src/qemu/MIGRATION.txt +++ b/src/qemu/MIGRATION.txt @@ -74,7 +74,7 @@ The sequence of calling qemuMigrationJob* helper methods is as follows: migration type and version) has to start migration job and keep it active: qemuMigrationJobStart(driver, vm, QEMU_JOB_MIGRATION_{IN,OUT}); - qemuMigrationJobSetPhase(driver, vm, QEMU_MIGRATION_PHASE_*); + qemuMigrationJobSetPhase(driver, vm, VIR_MIGRATION_PHASE_*); ...do work... qemuMigrationJobContinue(vm); @@ -82,7 +82,7 @@ The sequence of calling qemuMigrationJob* helper methods is as follows: if (!qemuMigrationJobIsActive(vm, QEMU_JOB_MIGRATION_{IN,OUT})) return; - qemuMigrationJobStartPhase(driver, vm, QEMU_MIGRATION_PHASE_*); + qemuMigrationJobStartPhase(driver, vm, VIR_MIGRATION_PHASE_*); ...do work... qemuMigrationJobContinue(vm); @@ -90,11 +90,11 @@ The sequence of calling qemuMigrationJob* helper methods is as follows: if (!qemuMigrationJobIsActive(vm, QEMU_JOB_MIGRATION_{IN,OUT})) return; - qemuMigrationJobStartPhase(driver, vm, QEMU_MIGRATION_PHASE_*); + qemuMigrationJobStartPhase(driver, vm, VIR_MIGRATION_PHASE_*); ...do work... qemuMigrationJobFinish(driver, vm); While migration job is running (i.e., after qemuMigrationJobStart* but before qemuMigrationJob{Continue,Finish}), migration phase can be advanced using - qemuMigrationJobSetPhase(driver, vm, QEMU_MIGRATION_PHASE_*); + qemuMigrationJobSetPhase(driver, vm, VIR_MIGRATION_PHASE_*); diff --git a/src/qemu/qemu_domainjob.c b/src/qemu/qemu_domainjob.c index eeaa089959..02f85e1b13 100644 --- a/src/qemu/qemu_domainjob.c +++ b/src/qemu/qemu_domainjob.c @@ -70,7 +70,7 @@ qemuDomainAsyncJobPhaseToString(qemuDomainAsyncJob job, switch (job) { case QEMU_ASYNC_JOB_MIGRATION_OUT: case QEMU_ASYNC_JOB_MIGRATION_IN: - return qemuMigrationJobPhaseTypeToString(phase); + return virMigrationJobPhaseTypeToString(phase); case QEMU_ASYNC_JOB_SAVE: case QEMU_ASYNC_JOB_DUMP: @@ -96,7 +96,7 @@ qemuDomainAsyncJobPhaseFromString(qemuDomainAsyncJob job, switch (job) { case QEMU_ASYNC_JOB_MIGRATION_OUT: case QEMU_ASYNC_JOB_MIGRATION_IN: - return qemuMigrationJobPhaseTypeFromString(phase); + return virMigrationJobPhaseTypeFromString(phase); case QEMU_ASYNC_JOB_SAVE: case QEMU_ASYNC_JOB_DUMP: diff --git a/src/qemu/qemu_migration.c b/src/qemu/qemu_migration.c index 142faa2cf9..a6f1da97bd 100644 --- a/src/qemu/qemu_migration.c +++ b/src/qemu/qemu_migration.c @@ -67,8 +67,8 @@ VIR_LOG_INIT("qemu.qemu_migration"); -VIR_ENUM_IMPL(qemuMigrationJobPhase, - QEMU_MIGRATION_PHASE_LAST, +VIR_ENUM_IMPL(virMigrationJobPhase, + VIR_MIGRATION_PHASE_LAST, "none", "perform2", "begin3", @@ -91,13 +91,13 @@ qemuMigrationJobStart(virQEMUDriverPtr driver, static void qemuMigrationJobSetPhase(virQEMUDriverPtr driver, virDomainObjPtr vm, - qemuMigrationJobPhase phase) + virMigrationJobPhase phase) ATTRIBUTE_NONNULL(1) ATTRIBUTE_NONNULL(2); static void qemuMigrationJobStartPhase(virQEMUDriverPtr driver, virDomainObjPtr vm, - qemuMigrationJobPhase phase) + virMigrationJobPhase phase) ATTRIBUTE_NONNULL(1) ATTRIBUTE_NONNULL(2); static void @@ -2023,13 +2023,13 @@ qemuMigrationSrcCleanup(virDomainObjPtr vm, " was closed; canceling the migration", vm->def->name); - switch ((qemuMigrationJobPhase) priv->job.phase) { - case QEMU_MIGRATION_PHASE_BEGIN3: + switch ((virMigrationJobPhase) priv->job.phase) { + case VIR_MIGRATION_PHASE_BEGIN3: /* just forget we were about to migrate */ qemuDomainObjDiscardAsyncJob(driver, vm); break; - case QEMU_MIGRATION_PHASE_PERFORM3_DONE: + case VIR_MIGRATION_PHASE_PERFORM3_DONE: VIR_WARN("Migration of domain %s finished but we don't know if the" " domain was successfully started on destination or not", vm->def->name); @@ -2039,19 +2039,19 @@ qemuMigrationSrcCleanup(virDomainObjPtr vm, qemuDomainObjDiscardAsyncJob(driver, vm); break; - case QEMU_MIGRATION_PHASE_PERFORM3: + case VIR_MIGRATION_PHASE_PERFORM3: /* cannot be seen without an active migration API; unreachable */ - case QEMU_MIGRATION_PHASE_CONFIRM3: - case QEMU_MIGRATION_PHASE_CONFIRM3_CANCELLED: + case VIR_MIGRATION_PHASE_CONFIRM3: + case VIR_MIGRATION_PHASE_CONFIRM3_CANCELLED: /* all done; unreachable */ - case QEMU_MIGRATION_PHASE_PREPARE: - case QEMU_MIGRATION_PHASE_FINISH2: - case QEMU_MIGRATION_PHASE_FINISH3: + case VIR_MIGRATION_PHASE_PREPARE: + case VIR_MIGRATION_PHASE_FINISH2: + case VIR_MIGRATION_PHASE_FINISH3: /* incoming migration; unreachable */ - case QEMU_MIGRATION_PHASE_PERFORM2: + case VIR_MIGRATION_PHASE_PERFORM2: /* single phase outgoing migration; unreachable */ - case QEMU_MIGRATION_PHASE_NONE: - case QEMU_MIGRATION_PHASE_LAST: + case VIR_MIGRATION_PHASE_NONE: + case VIR_MIGRATION_PHASE_LAST: /* unreachable */ ; } @@ -2087,7 +2087,7 @@ qemuMigrationSrcBeginPhase(virQEMUDriverPtr driver, * change protection. */ if (priv->job.asyncJob == QEMU_ASYNC_JOB_MIGRATION_OUT) - qemuMigrationJobSetPhase(driver, vm, QEMU_MIGRATION_PHASE_BEGIN3); + qemuMigrationJobSetPhase(driver, vm, VIR_MIGRATION_PHASE_BEGIN3); if (!qemuMigrationSrcIsAllowed(driver, vm, true, flags)) return NULL; @@ -2546,7 +2546,7 @@ qemuMigrationDstPrepareAny(virQEMUDriverPtr driver, if (qemuMigrationJobStart(driver, vm, QEMU_ASYNC_JOB_MIGRATION_IN, flags) < 0) goto cleanup; - qemuMigrationJobSetPhase(driver, vm, QEMU_MIGRATION_PHASE_PREPARE); + qemuMigrationJobSetPhase(driver, vm, VIR_MIGRATION_PHASE_PREPARE); /* Domain starts inactive, even if the domain XML had an id field. */ vm->def->id = -1; @@ -3007,10 +3007,9 @@ qemuMigrationSrcConfirmPhase(virQEMUDriverPtr driver, virCheckFlags(QEMU_MIGRATION_FLAGS, -1); - qemuMigrationJobSetPhase(driver, vm, - retcode == 0 - ? QEMU_MIGRATION_PHASE_CONFIRM3 - : QEMU_MIGRATION_PHASE_CONFIRM3_CANCELLED); + qemuMigrationJobSetPhase(driver, vm, retcode == 0 + ? VIR_MIGRATION_PHASE_CONFIRM3 + : VIR_MIGRATION_PHASE_CONFIRM3_CANCELLED); if (!(mig = qemuMigrationEatCookie(driver, vm->def, priv->origname, priv, cookiein, cookieinlen, @@ -3100,7 +3099,7 @@ qemuMigrationSrcConfirm(virQEMUDriverPtr driver, unsigned int flags, int cancelled) { - qemuMigrationJobPhase phase; + virMigrationJobPhase phase; virQEMUDriverConfigPtr cfg = NULL; int ret = -1; @@ -3110,9 +3109,9 @@ qemuMigrationSrcConfirm(virQEMUDriverPtr driver, goto cleanup; if (cancelled) - phase = QEMU_MIGRATION_PHASE_CONFIRM3_CANCELLED; + phase = VIR_MIGRATION_PHASE_CONFIRM3_CANCELLED; else - phase = QEMU_MIGRATION_PHASE_CONFIRM3; + phase = VIR_MIGRATION_PHASE_CONFIRM3; qemuMigrationJobStartPhase(driver, vm, phase); virCloseCallbacksUnset(driver->closeCallbacks, vm, @@ -4059,7 +4058,7 @@ qemuMigrationSrcPerformPeer2Peer2(virQEMUDriverPtr driver, * until the migration is complete. */ VIR_DEBUG("Perform %p", sconn); - qemuMigrationJobSetPhase(driver, vm, QEMU_MIGRATION_PHASE_PERFORM2); + qemuMigrationJobSetPhase(driver, vm, VIR_MIGRATION_PHASE_PERFORM2); if (flags & VIR_MIGRATE_TUNNELLED) ret = qemuMigrationSrcPerformTunnel(driver, vm, st, NULL, NULL, 0, NULL, NULL, @@ -4297,7 +4296,7 @@ qemuMigrationSrcPerformPeer2Peer3(virQEMUDriverPtr driver, * confirm migration completion. */ VIR_DEBUG("Perform3 %p uri=%s", sconn, NULLSTR(uri)); - qemuMigrationJobSetPhase(driver, vm, QEMU_MIGRATION_PHASE_PERFORM3); + qemuMigrationJobSetPhase(driver, vm, VIR_MIGRATION_PHASE_PERFORM3); VIR_FREE(cookiein); cookiein = g_steal_pointer(&cookieout); cookieinlen = cookieoutlen; @@ -4322,8 +4321,7 @@ qemuMigrationSrcPerformPeer2Peer3(virQEMUDriverPtr driver, if (ret < 0) { virErrorPreserveLast(&orig_err); } else { - qemuMigrationJobSetPhase(driver, vm, - QEMU_MIGRATION_PHASE_PERFORM3_DONE); + qemuMigrationJobSetPhase(driver, vm, VIR_MIGRATION_PHASE_PERFORM3_DONE); } /* If Perform returns < 0, then we need to cancel the VM @@ -4687,7 +4685,7 @@ qemuMigrationSrcPerformJob(virQEMUDriverPtr driver, migParams, flags, dname, resource, &v3proto); } else { - qemuMigrationJobSetPhase(driver, vm, QEMU_MIGRATION_PHASE_PERFORM2); + qemuMigrationJobSetPhase(driver, vm, VIR_MIGRATION_PHASE_PERFORM2); ret = qemuMigrationSrcPerformNative(driver, vm, persist_xml, uri, cookiein, cookieinlen, cookieout, cookieoutlen, flags, resource, NULL, NULL, 0, NULL, @@ -4772,7 +4770,7 @@ qemuMigrationSrcPerformPhase(virQEMUDriverPtr driver, return ret; } - qemuMigrationJobStartPhase(driver, vm, QEMU_MIGRATION_PHASE_PERFORM3); + qemuMigrationJobStartPhase(driver, vm, VIR_MIGRATION_PHASE_PERFORM3); virCloseCallbacksUnset(driver->closeCallbacks, vm, qemuMigrationSrcCleanup); @@ -4786,7 +4784,7 @@ qemuMigrationSrcPerformPhase(virQEMUDriverPtr driver, goto endjob; } - qemuMigrationJobSetPhase(driver, vm, QEMU_MIGRATION_PHASE_PERFORM3_DONE); + qemuMigrationJobSetPhase(driver, vm, VIR_MIGRATION_PHASE_PERFORM3_DONE); if (virCloseCallbacksSet(driver->closeCallbacks, vm, conn, qemuMigrationSrcCleanup) < 0) @@ -5019,8 +5017,8 @@ qemuMigrationDstFinish(virQEMUDriverPtr driver, ignore_value(virTimeMillisNow(&timeReceived)); qemuMigrationJobStartPhase(driver, vm, - v3proto ? QEMU_MIGRATION_PHASE_FINISH3 - : QEMU_MIGRATION_PHASE_FINISH2); + v3proto ? VIR_MIGRATION_PHASE_FINISH3 + : VIR_MIGRATION_PHASE_FINISH2); qemuDomainCleanupRemove(vm, qemuMigrationDstPrepareCleanup); g_clear_pointer(&priv->job.completed, qemuDomainJobInfoFree); @@ -5498,14 +5496,14 @@ qemuMigrationJobStart(virQEMUDriverPtr driver, static void qemuMigrationJobSetPhase(virQEMUDriverPtr driver, virDomainObjPtr vm, - qemuMigrationJobPhase phase) + virMigrationJobPhase phase) { qemuDomainObjPrivatePtr priv = vm->privateData; if (phase < priv->job.phase) { VIR_ERROR(_("migration protocol going backwards %s => %s"), - qemuMigrationJobPhaseTypeToString(priv->job.phase), - qemuMigrationJobPhaseTypeToString(phase)); + virMigrationJobPhaseTypeToString(priv->job.phase), + virMigrationJobPhaseTypeToString(phase)); return; } @@ -5515,7 +5513,7 @@ qemuMigrationJobSetPhase(virQEMUDriverPtr driver, static void qemuMigrationJobStartPhase(virQEMUDriverPtr driver, virDomainObjPtr vm, - qemuMigrationJobPhase phase) + virMigrationJobPhase phase) { qemuMigrationJobSetPhase(driver, vm, phase); } diff --git a/src/qemu/qemu_migration.h b/src/qemu/qemu_migration.h index b6f88d3fd9..b05f5254b4 100644 --- a/src/qemu/qemu_migration.h +++ b/src/qemu/qemu_migration.h @@ -24,6 +24,7 @@ #include "qemu_conf.h" #include "qemu_domain.h" #include "qemu_migration_params.h" +#include "virmigration.h" #include "virenum.h" /* @@ -87,22 +88,6 @@ NULL -typedef enum { - QEMU_MIGRATION_PHASE_NONE = 0, - QEMU_MIGRATION_PHASE_PERFORM2, - QEMU_MIGRATION_PHASE_BEGIN3, - QEMU_MIGRATION_PHASE_PERFORM3, - QEMU_MIGRATION_PHASE_PERFORM3_DONE, - QEMU_MIGRATION_PHASE_CONFIRM3_CANCELLED, - QEMU_MIGRATION_PHASE_CONFIRM3, - QEMU_MIGRATION_PHASE_PREPARE, - QEMU_MIGRATION_PHASE_FINISH2, - QEMU_MIGRATION_PHASE_FINISH3, - - QEMU_MIGRATION_PHASE_LAST -} qemuMigrationJobPhase; -VIR_ENUM_DECL(qemuMigrationJobPhase); - char * qemuMigrationSrcBegin(virConnectPtr conn, virDomainObjPtr vm, diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index ad461d8f34..55b31382f3 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -3434,24 +3434,24 @@ qemuProcessRecoverMigrationIn(virQEMUDriverPtr driver, (state == VIR_DOMAIN_RUNNING && reason == VIR_DOMAIN_RUNNING_POSTCOPY); - switch ((qemuMigrationJobPhase) job->phase) { - case QEMU_MIGRATION_PHASE_NONE: - case QEMU_MIGRATION_PHASE_PERFORM2: - case QEMU_MIGRATION_PHASE_BEGIN3: - case QEMU_MIGRATION_PHASE_PERFORM3: - case QEMU_MIGRATION_PHASE_PERFORM3_DONE: - case QEMU_MIGRATION_PHASE_CONFIRM3_CANCELLED: - case QEMU_MIGRATION_PHASE_CONFIRM3: - case QEMU_MIGRATION_PHASE_LAST: + switch ((virMigrationJobPhase) job->phase) { + case VIR_MIGRATION_PHASE_NONE: + case VIR_MIGRATION_PHASE_PERFORM2: + case VIR_MIGRATION_PHASE_BEGIN3: + case VIR_MIGRATION_PHASE_PERFORM3: + case VIR_MIGRATION_PHASE_PERFORM3_DONE: + case VIR_MIGRATION_PHASE_CONFIRM3_CANCELLED: + case VIR_MIGRATION_PHASE_CONFIRM3: + case VIR_MIGRATION_PHASE_LAST: /* N/A for incoming migration */ break; - case QEMU_MIGRATION_PHASE_PREPARE: + case VIR_MIGRATION_PHASE_PREPARE: VIR_DEBUG("Killing unfinished incoming migration for domain %s", vm->def->name); return -1; - case QEMU_MIGRATION_PHASE_FINISH2: + case VIR_MIGRATION_PHASE_FINISH2: /* source domain is already killed so let's just resume the domain * and hope we are all set */ VIR_DEBUG("Incoming migration finished, resuming domain %s", @@ -3463,7 +3463,7 @@ qemuProcessRecoverMigrationIn(virQEMUDriverPtr driver, } break; - case QEMU_MIGRATION_PHASE_FINISH3: + case VIR_MIGRATION_PHASE_FINISH3: /* migration finished, we started resuming the domain but didn't * confirm success or failure yet; killing it seems safest unless * we already started guest CPUs or we were in post-copy mode */ @@ -3495,22 +3495,22 @@ qemuProcessRecoverMigrationOut(virQEMUDriverPtr driver, reason == VIR_DOMAIN_PAUSED_POSTCOPY_FAILED); bool resume = false; - switch ((qemuMigrationJobPhase) job->phase) { - case QEMU_MIGRATION_PHASE_NONE: - case QEMU_MIGRATION_PHASE_PREPARE: - case QEMU_MIGRATION_PHASE_FINISH2: - case QEMU_MIGRATION_PHASE_FINISH3: - case QEMU_MIGRATION_PHASE_LAST: + switch ((virMigrationJobPhase) job->phase) { + case VIR_MIGRATION_PHASE_NONE: + case VIR_MIGRATION_PHASE_PREPARE: + case VIR_MIGRATION_PHASE_FINISH2: + case VIR_MIGRATION_PHASE_FINISH3: + case VIR_MIGRATION_PHASE_LAST: /* N/A for outgoing migration */ break; - case QEMU_MIGRATION_PHASE_BEGIN3: + case VIR_MIGRATION_PHASE_BEGIN3: /* nothing happened so far, just forget we were about to migrate the * domain */ break; - case QEMU_MIGRATION_PHASE_PERFORM2: - case QEMU_MIGRATION_PHASE_PERFORM3: + case VIR_MIGRATION_PHASE_PERFORM2: + case VIR_MIGRATION_PHASE_PERFORM3: /* migration is still in progress, let's cancel it and resume the * domain; however we can only do that before migration enters * post-copy mode @@ -3528,7 +3528,7 @@ qemuProcessRecoverMigrationOut(virQEMUDriverPtr driver, } break; - case QEMU_MIGRATION_PHASE_PERFORM3_DONE: + case VIR_MIGRATION_PHASE_PERFORM3_DONE: /* migration finished but we didn't have a chance to get the result * of Finish3 step; third party needs to check what to do next; in * post-copy mode we can use PAUSED_POSTCOPY_FAILED state for this @@ -3537,7 +3537,7 @@ qemuProcessRecoverMigrationOut(virQEMUDriverPtr driver, qemuMigrationAnyPostcopyFailed(driver, vm); break; - case QEMU_MIGRATION_PHASE_CONFIRM3_CANCELLED: + case VIR_MIGRATION_PHASE_CONFIRM3_CANCELLED: /* Finish3 failed, we need to resume the domain, but once we enter * post-copy mode there's no way back, so let's just mark the domain * as broken in that case @@ -3551,7 +3551,7 @@ qemuProcessRecoverMigrationOut(virQEMUDriverPtr driver, } break; - case QEMU_MIGRATION_PHASE_CONFIRM3: + case VIR_MIGRATION_PHASE_CONFIRM3: /* migration completed, we need to kill the domain here */ *stopFlags |= VIR_QEMU_PROCESS_STOP_MIGRATED; return -1; -- 2.25.1

2 1

[GSoC][PATCH v4 0/8] creating hypervisor-agnostic domainjob
by Prathamesh Chavan 26 Aug '20

26 Aug '20

This series deals with removal of external dependencies of `qemu_domainjob`, followed by it being moved to `hypervisor/virdomainjob`, and renaming corresponding structures and functions for creating the hypervisor-agnostic domain-jobs. Previous series can be seen here[1]. The previous series was updated because: 1. The patch series was rebased on master, after the recent merge of `qemu_snapshot`, so that while checking it out, we can avoid conflicts. 2. Patch "qemu_domainjob: removed reference to `qemuDomainObjPrivatePtr`" was updated. 3. Patch "virdomainjob: moved `qemu_domainjob` to `hypervisor/virdomainjob`" was added to complete the series. This series is based on this patch (virmigraiton: `qemuMigrationJobPhase` transformed for more generic use)[2]. Please checkout that out first and then base this patch series over it. You can see all these changes on my github account[3]. [1]: https://www.redhat.com/archives/libvir-list/2020-August/msg00744.html [2]: https://www.redhat.com/archives/libvir-list/2020-August/msg00966.html [3]: https://github.com/pratham-pc/libvirt/tree/cms Prathamesh Chavan (8): qemu_domainjob: `qemuDomainJobInfo` code moved to `qemu_domain` qemu_domain: Added `qemuDomainJobInfo` to domainJob's `privateData` qemu_domainjob: callback functions added to access `jobs_queued` qemu_domainjob: callback function added to access `maxQueuedJobs` qemu_domainjob: `qemuDomainJobPrivateJobCallbacks` structure nested qemu_domainjob: add `saveDomainStatus` as a callback function to jobs qemu_domainjob: removed reference to `qemuDomainObjPrivatePtr` virdomainjob: moved `qemu_domainjob` to `hypervisor/virdomainjob` po/POTFILES.in | 2 +- po/libvirt.pot | 34 +- src/hypervisor/meson.build | 1 + src/hypervisor/virdomainjob.c | 792 ++++++++++++++++++ src/hypervisor/virdomainjob.h | 243 ++++++ src/libvirt_private.syms | 28 + src/qemu/meson.build | 1 - src/qemu/qemu_backup.c | 92 +-- src/qemu/qemu_backup.h | 5 +- src/qemu/qemu_block.c | 65 +- src/qemu/qemu_block.h | 18 +- src/qemu/qemu_blockjob.c | 77 +- src/qemu/qemu_blockjob.h | 3 +- src/qemu/qemu_checkpoint.c | 39 +- src/qemu/qemu_domain.c | 657 +++++++++++++-- src/qemu/qemu_domain.h | 106 ++- src/qemu/qemu_domainjob.c | 1291 ------------------------------ src/qemu/qemu_domainjob.h | 291 ------- src/qemu/qemu_driver.c | 1186 ++++++++++++++------------- src/qemu/qemu_hotplug.c | 335 ++++---- src/qemu/qemu_hotplug.h | 38 +- src/qemu/qemu_migration.c | 594 +++++++------- src/qemu/qemu_migration.h | 20 +- src/qemu/qemu_migration_cookie.c | 17 +- src/qemu/qemu_migration_params.c | 52 +- src/qemu/qemu_migration_params.h | 15 +- src/qemu/qemu_process.c | 462 ++++++----- src/qemu/qemu_process.h | 37 +- src/qemu/qemu_saveimage.c | 4 +- src/qemu/qemu_saveimage.h | 6 +- src/qemu/qemu_snapshot.c | 111 ++- tests/qemuhotplugtest.c | 2 +- 32 files changed, 3323 insertions(+), 3301 deletions(-) create mode 100644 src/hypervisor/virdomainjob.c create mode 100644 src/hypervisor/virdomainjob.h delete mode 100644 src/qemu/qemu_domainjob.c delete mode 100644 src/qemu/qemu_domainjob.h -- 2.25.1

1 8

[PATCH 0/3] CI/test improvements in qemuxml2argvtest
by Daniel Henrique Barboza 25 Aug '20

25 Aug '20

Hi, This started as a saga of figuring out why Libvirt was failing to build in a Power 9 host after the switch to meson. Turns out that one of the tests, qemuxml2argvtest, is taking almost 5 minutes to complete, and meson throws a timeout error. This wasn't noticed before because the test actually completes without errors. Running 'perf' showed that 95% of the time spent in qemuxml2argvtest in Power 9 were in virHostCPUGetMicrocodeVersion() calls, which is not needed for PowerPC since 'microcode' is x86 specific. This means that we're calling 'show_cpuinfo' from the kernel, fetching dynamic data from all CPUs, and then failing because there is no 'microcode' string. The Power 9 host I tested with has 128 CPUs, meaning that 'show_cpuinfo' is not something to scoff at. First patch is a trivial cleanup. Patch 02 is a runtime fix. The cake is in patch 03. Patch 03 will improve CI on all archs, including x86, which is nice. Daniel Henrique Barboza (3): virhostcpu.c: modernize virHostCPUGetMicrocodeVersion() virhostcpu.c: skip non x86 hosts in virHostCPUGetMicrocodeVersion() domaincapsmoc.c: mock virHostCPUGetMicrocodeVersion() src/cpu/cpu_x86.c | 2 +- src/qemu/qemu_capabilities.c | 4 ++-- src/util/virhostcpu.c | 13 +++++++------ src/util/virhostcpu.h | 3 ++- tests/domaincapsmock.c | 6 ++++++ 5 files changed, 18 insertions(+), 10 deletions(-) -- 2.26.2

2 8

[PATCH V2 0/3] Xen: Add support for qemu commandline passthrough
by Jim Fehlig 25 Aug '20

25 Aug '20

V2 of https://www.redhat.com/archives/libvir-list/2020-July/msg02043.html Changes in V2: - Drop use of virStringListCopy, and hence patch 1/4 - Switch to using 'xen' namespace and schema extenstion instead of qemu - Use more glib functions Jim Fehlig (3): Xen: Add support for qemu command-line passthrough Xen: Add support for qemu commandline passthrough to config converter news: Mention support for device model commandline passthrough in Xen NEWS.rst | 8 ++ docs/drvxen.html.in | 29 +++++++ docs/schemas/domaincommon.rng | 17 ++++ src/libxl/libxl_conf.c | 10 ++- src/libxl/libxl_conf.h | 8 ++ src/libxl/libxl_domain.c | 84 ++++++++++++++++++ src/libxl/libxl_domain.h | 1 + src/libxl/xen_xl.c | 91 ++++++++++++++++++++ tests/xlconfigdata/test-qemu-passthrough.cfg | 26 ++++++ tests/xlconfigdata/test-qemu-passthrough.xml | 53 ++++++++++++ tests/xlconfigtest.c | 1 + 11 files changed, 327 insertions(+), 1 deletion(-) create mode 100644 tests/xlconfigdata/test-qemu-passthrough.cfg create mode 100644 tests/xlconfigdata/test-qemu-passthrough.xml -- 2.26.2

2 7

XDR related breakage in libvirt v6.6.0 when using libvirt-lxc
by Christian Ehrhardt 25 Aug '20

25 Aug '20

Hi, I expect that this falls under the "with meson now everything is different anyway" umbrella but wanted to let you know about this as it affects v6.6 in at least Ubuntu/Debian. The following recent patch has broken libvirt-lxc for us: commit d7147b3797380de2d159ce6324536f3e1f2d97e3 Author: Pavel Hrdina <phrdina(a)redhat.com> Date: Fri Jun 19 00:44:07 2020 +0200 m4: virt-xdr: rewrite XDR check I was tracking that down for [1] since the tests [4] failed on me. [2] holds the backtrace. In Debian the tests are skipped which explains why they were not seen there: smoke-lxc SKIP Test requires machine-level isolation but testbed does not provide that What happens is that the libvirt_lxc segfaults when using XDR functions. dmesg shows: [582093.524644] libvirt_lxc[261446]: segfault at 0 ip 0000000000000000 sp 00007ffdd2345598 error 14 in libvirt_lxc[5587e42aa000+8000] [582093.524650] Code: Bad RIP value. There are quite some uncertainties left, but on the surface it seems that it links with libtirpc but then instead of calling libtirpc: src/xdr.c:929:xdr_uint64_t(xdrs, ullp) it ends (gdb tells us in [2]) in glibc glibc: sunrpc/xdr_intXX_t.c:62:xdr_uint64_t (XDR *xdrs, uint64_t *uip) And the return from that function breaks it badly (instruction pointer at 0x0 -> segfault) Bisecting pointed to the referred commit which brings libtirpc into the mix. The former builds had xdr detected, but not with libtirpc. configure: xdr: yes (CFLAGS='' LIBS='') The new config now does configure: xdr: yes (CFLAGS='-I/usr/include/tirpc' LIBS='-ltirpc') And the resulting libvirt_lxc reflects that v6.0.0 $ lddtree /usr/lib/libvirt/libvirt_lxc | grep tirpc v6.6.0 $ lddtree /usr/lib/libvirt/libvirt_lxc | grep tirpc libtirpc.so.3 => /lib/x86_64-linux-gnu/libtirpc.so.3 This seems to lead to the bad jump and the crash eventually. Meanwhile reverting d7147b37 "m4: virt-xdr: rewrite XDR check" on top of v6.6.0 resolves the issue back to the former state. For anyone that wants to recreate this, I also attached a bisect script [3] which includes the test case you'd need. [1]: https://bugs.launchpad.net/ubuntu/+source/libvirt/+bug/1892826 [2]: https://bugs.launchpad.net/ubuntu/+source/libvirt/+bug/1892826/comments/4 [3]: https://bugs.launchpad.net/ubuntu/+source/libvirt/+bug/1892826/+attachment/… [4]: https://objectstorage.prodstack4-5.canonical.com/v1/AUTH_77e2ada1e7a84929a7… -- Christian Ehrhardt Staff Engineer, Ubuntu Server Canonical Ltd

2 3

[libvirt PATCH] storage_util: fix qemu-img sparse allocation
by Pavel Hrdina 25 Aug '20

25 Aug '20

Commit <c9ec7088c7a3f4cd26bb471f1f243931fff6f4f9> introduced a support to fully allocate qcow2 images when <allocation> matches <capacity> but it doesn't work as expected. The issue is that info.size_arg is in KB but the info.allocation introduced by the mentioned commit is in B. This results in using "preallocation=falloc," in cases where "preallocation=metadata," should be used. Signed-off-by: Pavel Hrdina <phrdina(a)redhat.com> --- src/storage/storage_util.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/storage/storage_util.c b/src/storage/storage_util.c index f7c09e3375..fcecedbc3a 100644 --- a/src/storage/storage_util.c +++ b/src/storage/storage_util.c @@ -1044,7 +1044,7 @@ virStorageBackendCreateQemuImgCmdFromVol(virStoragePoolObjPtr pool, .type = NULL, .inputType = NULL, .path = vol->target.path, - .allocation = vol->target.allocation, + .allocation = VIR_DIV_UP(vol->target.allocation, 1024), .encryption = !!vol->target.encryption, .preallocate = !!(flags & VIR_STORAGE_VOL_CREATE_PREALLOC_METADATA), .compat = vol->target.compat, -- 2.26.2

2 1