- Devel - libvirt List Archives

[PATCH RFC v3 00/14] Add riscv kvm accel support
by Yifei Jiang 27 Aug '20

27 Aug '20

This series adds both riscv32 and riscv64 kvm support, and implements migration based on riscv. It is based on temporarily unaccepted kvm: https://github.com/kvm-riscv/linux Compared to RFC v2, the migration is supported in this series. The new added migration feature also requires the following patches: [1] [PATCH RFC 0/2] Add log dirty support https://www.mail-archive.com/linux-kernel@vger.kernel.org/msg2284945.html [2] [PATCH] target/riscv: raise exception to HS-mode at get_physical_address https://patchew.org/QEMU/20200824084158.1769-1-jiangyifei@huawei.com/ Several steps to use this: 1. Build emulation $ ./configure --target-list=riscv64-softmmu $ make -j$(nproc) 2. Build kernel https://github.com/kvm-riscv/linux 3. Build QEMU VM I cross built in riscv toolchain. $ PKG_CONFIG_LIBDIR=<toolchain pkgconfig path> $ export PKG_CONFIG_SYSROOT_DIR=<toolchain sysroot path> $ ./configure --target-list=riscv64-softmmu --enable-kvm \ --cross-prefix=riscv64-linux-gnu- --disable-libiscsi --disable-glusterfs \ --disable-libusb --disable-usb-redir --audio-drv-list= --disable-opengl \ --disable-libxml2 $ make -j$(nproc) 4. Start emulation $ ./qemu-system-riscv64 -M virt -m 4096M -cpu rv64,x-h=true -nographic \ -name guest=riscv-hyp,debug-threads=on \ -smp 4 \ -bios ./fw_jump.bin \ -kernel ./Image \ -drive file=./hyp.img,format=raw,id=hd0 \ -device virtio-blk-device,drive=hd0 \ -append "root=/dev/vda rw console=ttyS0 earlycon=sbi" 5. Start kvm-acceled QEMU VM in emulation $ ./qemu-system-riscv64 -M virt,accel=kvm -m 1024M -cpu host -nographic \ -name guest=riscv-guset \ -smp 2 \ -bios none \ -kernel ./Image \ -drive file=./guest.img,format=raw,id=hd0 \ -device virtio-blk-device,drive=hd0 \ -append "root=/dev/vda rw console=ttyS0 earlycon=sbi" This following link describes the details about live migration steps: https://gitee.com/openeuler/qemu/wikis/Live%20migration?sort_id=2767831 Changes since RFC v2 1. Fix checkpatch error at target/riscv/sbi_ecall_interface.h. 2. Add riscv migration support. Changes since RFC v1 1. Add separate SBI ecall interface header. 2. Add riscv32 kvm accel support. Yifei Jiang (14): linux-header: Update linux/kvm.h target/riscv: Add target/riscv/kvm.c to place the public kvm interface target/riscv: Implement function kvm_arch_init_vcpu target/riscv: Implement kvm_arch_get_registers arget/riscv: Implement kvm_arch_put_registers target/riscv: Support start kernel directly by KVM hw/riscv: PLIC update external interrupt by KVM when kvm enabled target/riscv: Handler KVM_EXIT_RISCV_SBI exit target/riscv: Add host cpu type target/riscv: Add sifive_plic vmstate target/riscv: Support riscv cpu vmstate target/riscv: Add kvm_riscv_get/put_regs_timer target/riscv: Implement virtual time adjusting with vm state changing target/riscv: Support virtual time context synchronization configure | 1 + hw/riscv/sifive_plic.c | 55 ++- hw/riscv/virt.c | 8 + include/hw/riscv/sifive_plic.h | 1 + linux-headers/linux/kvm.h | 8 + target/riscv/cpu.c | 41 +- target/riscv/cpu.h | 10 + target/riscv/kvm.c | 599 +++++++++++++++++++++++++++++ target/riscv/kvm_riscv.h | 25 ++ target/riscv/meson.build | 1 + target/riscv/sbi_ecall_interface.h | 72 ++++ 11 files changed, 809 insertions(+), 12 deletions(-) create mode 100644 target/riscv/kvm.c create mode 100644 target/riscv/kvm_riscv.h create mode 100644 target/riscv/sbi_ecall_interface.h -- 2.19.1

1 14

[libvirt PATCH v2 0/2] Eliminate old tap/macvtap teardown stomping on new tap setup
by Laine Stump 27 Aug '20

27 Aug '20

V1 is here: https://www.redhat.com/archives/libvir-list/2020-August/msg00756.html The problem and this solution are very well described in patches 2 and 3, but in short - because we (libvirt for macvtap, the kernel for tap) always try to assign the lowest numbered names possible to macvtap and tap devices, we sometimes create a new tap for a new guest using the same name as an old tap for an old guest that is shutting down simultaneous to setting up the new guest/tap. This can lead to the old guest teardown stomping on the new guest setup. This is the problem that the authors were attempting to solve in these two patches sent earlier in the summer: https://www.redhat.com/archives/libvir-list/2020-June/msg00481.html https://www.redhat.com/archives/libvir-list/2020-June/msg00525.html and also in this V2 patch, which Bingsong Si sent in response to my poorly-thought-out advice in my response to his original patch: https://www.redhat.com/archives/libvir-list/2020-June/msg00755.html Somewhere during that discussion, danpb suggested that in order to *really* solve the problem, we should use our own counter for auto-generated tap device names (instead of relying on the kernel) and just never re-use a name until the counter rolls over. That's essentially what these two patches do. One possibly undesirable side effect of this (and the other) patch is that the longer a host is running without reboot, the higher the numbers tap device names will get. While users are accustomed to always seeing vnet0 and vnet1, they may be a bit surprised to now see vnet39283 or macvtap735. It has been pointed out to me (again by danpb) that the same thing happened with PIDs a few years ago, and while it looked strange at first, everyone is now accustomed to it. Changes from V1: Patch 1 from V1 was removed - everything it changed is now removed/replaced in the new Patch 1 (which was Patch 2 in V1). And so of course, what was Patch 3 in V1 is now Patch 2 in V2. I eliminated the old bitmap reservation system in the macvtap patch (Patch 1) rather than adding on top of it as I had in V1 - it really was beyond redundant and unnecessary, and just clouded up the whole situation. This also allowed me to get rid of the 8192 limit (which was there only to limit the size of the virBitmap, which we no longer need), and allow the device names to count up until they overflow either the ifname[IFNAMSIZ] buffer, or reach INT_MAX. Likewise, I modified the standard tap patch to remove the artificial maximum of 99999, and just let it count up until it overflows. (Jano suggested that I should have a test case to test the entire range, but I don't think anyone would be happy with that. If I was masochistic and wanted to mock a bunch of virNetDev functions I could artificially test it by bumping up the counters with calls to the virNetDevTapReserveName() function, but it's 1AM. I did test the rollover of both cases (macvtap, where it overflows the buffer size first, and standard tap where it overflow the 32 bit int first) with a one-off build that started the counter just a few below the overflow point, and it does work correctly.) Laine Stump (2): util: replace macvtap name reservation bitmap with a simple counter util: assign tap device names using a monotonically increasing integer src/libvirt_private.syms | 2 +- src/libxl/libxl_driver.c | 2 +- src/lxc/lxc_process.c | 2 +- src/qemu/qemu_process.c | 22 +- src/util/virnetdevmacvlan.c | 402 +++++++++++++----------------------- src/util/virnetdevmacvlan.h | 6 +- src/util/virnetdevtap.c | 108 +++++++++- src/util/virnetdevtap.h | 4 + 8 files changed, 275 insertions(+), 273 deletions(-) -- 2.26.2

4 9

[PATCH] src/meson: add module name_suffix as 'so'
by Scott Shambarger 26 Aug '20

26 Aug '20

driver.c and locking/lock_manager.c both hardcode ".so" to module filenames. In meson, MacOS bundles (loadable objects) default to using ".dylib" - adding name_suffix as "so" will therefore align the code with all builds (no need to handle windows as it doesn't support libvirtd). Signed-off-by: Scott Shambarger <scott-libvirt(a)shambarger.net> --- src/meson.build | 1 + 1 file changed, 1 insertion(+) diff --git a/src/meson.build b/src/meson.build index 73ac99f01e..73986401c9 100644 --- a/src/meson.build +++ b/src/meson.build @@ -567,6 +567,7 @@ foreach module : virt_modules module['name'], module.get('sources', []), name_prefix: module.get('name_prefix', 'lib'), + name_suffix: 'so', include_directories: [ conf_inc_dir, module.get('include', []), -- 2.24.3 (Apple Git-128)

4 10

[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