December 2018 - Devel - libvirt List Archives

[libvirt] [PATCH] version: Add ParseVersion and a Version struct
by W. Trevor King 06 Dec '18

06 Dec '18

Make it easier to convert version integers to the more human-readable major.minor.release format. --- connect.go | 8 ++++++++ version.go | 52 ++++++++++++++++++++++++++++++++++++++++++++++ version_test.go | 64 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 124 insertions(+) create mode 100644 version.go create mode 100644 version_test.go diff --git a/connect.go b/connect.go index 8cc7cc7..fac45da 100644 --- a/connect.go +++ b/connect.go @@ -46,6 +46,7 @@ func init() { } const ( + // VERSION_NUMBER is the version of the linked libvirt. VERSION_NUMBER = uint32(C.LIBVIR_VERSION_NUMBER) ) @@ -306,7 +307,9 @@ func releaseConnectionData(c *Connect) { delete(connections, c.ptr) } +// GetVersion returns the version of the linked libvirt. // See also https://libvirt.org/html/libvirt-libvirt-host.html#virGetVersion +// and ParseVersion. func GetVersion() (uint32, error) { var version C.ulong var err C.virError @@ -540,7 +543,10 @@ func (c *Connect) GetHostname() (string, error) { return hostname, nil } +// GetLibVersion returns the version of libvirt used by the daemon +// running on the connected host. // See also https://libvirt.org/html/libvirt-libvirt-host.html#virConnectGetLibVersion +// and ParseVersion. func (c *Connect) GetLibVersion() (uint32, error) { var version C.ulong var err C.virError @@ -2272,7 +2278,9 @@ func (c *Connect) GetDomainCapabilities(emulatorbin string, arch string, machine return C.GoString(ret), nil } +// GetVersion returns the hypervisor version. // See also https://libvirt.org/html/libvirt-libvirt-host.html#virConnectGetVersion +// and ParseVersion. func (c *Connect) GetVersion() (uint32, error) { var hvVer C.ulong var err C.virError diff --git a/version.go b/version.go new file mode 100644 index 0000000..1a07fa3 --- /dev/null +++ b/version.go @@ -0,0 +1,52 @@ +/* + * This file is part of the libvirt-go project + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + * + * Copyright (C) 2018 Red Hat, Inc. + */ + +package libvirt + +import ( + "fmt" +) + +// Version represents the libvirt version. +// See also https://libvirt.org/html/libvirt-libvirt-host.html#virGetVersion +type Version struct { + Major uint32 + Minor uint32 + Release uint32 +} + +// String returns the "major.minor.release" version string. +func (v *Version) String() string { + return fmt.Sprintf("%d.%d.%d", v.Major, v.Minor, v.Release) +} + +// ParseVersion converts a version integer to a structured Version instance. +// See also https://libvirt.org/html/libvirt-libvirt-host.html#virGetVersion +func ParseVersion(version uint32) *Version { + return &Version{ + Major: version / 1000000, + Minor: (version / 1000) % 1000, + Release: version % 1000, + } +} diff --git a/version_test.go b/version_test.go new file mode 100644 index 0000000..ecafe5d --- /dev/null +++ b/version_test.go @@ -0,0 +1,64 @@ +/* + * This file is part of the libvirt-go project + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to deal + * in the Software without restriction, including without limitation the rights + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell + * copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN + * THE SOFTWARE. + * + * Copyright (C) 2018 Red Hat, Inc. + */ + +package libvirt + +import ( + "reflect" + "testing" +) + +func TestVersionString(t *testing.T) { + version := Version{ + Major: 3, + Minor: 9, + Release: 0, + } + actual := version.String() + expected := "3.9.0" + if actual != expected { + t.Fatalf("%q != %q", actual, expected) + } +} + +func TestParseVersion(t *testing.T) { + for _, testCase := range []struct{ + input uint32 + expected *Version + }{ + { + input: 3009000, + expected: &Version{Major: 3, Minor: 9}, + }, + { + input: 4001002, + expected: &Version{Major: 4, Minor: 1, Release: 2}, + }, + } { + actual := ParseVersion(testCase.input) + if !reflect.DeepEqual(actual, testCase.expected) { + t.Fatalf("%v != %v", actual, testCase.expected) + } + } +} -- 1.8.3.1

2 3

[libvirt] [PATCH 0/2] Fix a couple get all domain stats issues
by John Ferlan 06 Dec '18

06 Dec '18

One is longer term (patch1), while the other is sourced in this release (4.10.0) when IOThread stats were added. John Ferlan (2): qemu: Save qemuDomainGetStats error qemu: Don't fail stats collection due to IOThread capability src/qemu/qemu_driver.c | 28 ++++++++++++++++++++-------- 1 file changed, 20 insertions(+), 8 deletions(-) -- 2.17.2

4 13

[libvirt] [PATCH v2 00/18] Implement original label remembering
by Michal Privoznik 06 Dec '18

06 Dec '18

v2 of: https://www.redhat.com/archives/libvir-list/2018-November/msg00862.html diff to v1: - in 03/18 I've implemented FreeBSD support as discussed in v1 Michal Prívozník (18): security: Unify header conditionals util: Introduce xattr getter/setter/remover security: Include security_util security_dac: Restore label on failed chown() attempt virSecurityDACTransactionRun: Implement rollback virSecurityDACRestoreAllLabel: Reorder device relabeling virSecurityDACRestoreAllLabel: Restore more labels security_dac: Allow callers to enable/disable label remembering/recall security_dac: Remember old labels virSecurityDACRestoreImageLabelInt: Restore even shared/RO disks security_selinux: Track if transaction is restore security_selinux: Remember old labels security_selinux: Restore label on failed setfilecon() attempt virSecuritySELinuxTransactionRun: Implement rollback virSecuritySELinuxRestoreAllLabel: Reorder device relabeling virSecuritySELinuxRestoreAllLabel: Restore more labels tools: Provide a script to recover fubar'ed XATTRs setup qemu.conf: Allow users to enable/disable label remembering src/libvirt_private.syms | 3 + src/qemu/libvirtd_qemu.aug | 1 + src/qemu/qemu.conf | 6 + src/qemu/qemu_conf.c | 4 + src/qemu/test_libvirtd_qemu.aug.in | 1 + src/security/Makefile.inc.am | 2 + src/security/security_apparmor.h | 6 +- src/security/security_dac.c | 212 +++++++++++++++++------- src/security/security_dac.h | 6 +- src/security/security_driver.h | 6 +- src/security/security_manager.h | 6 +- src/security/security_nop.h | 6 +- src/security/security_selinux.c | 256 +++++++++++++++++++++-------- src/security/security_selinux.h | 6 +- src/security/security_stack.h | 6 +- src/security/security_util.c | 226 +++++++++++++++++++++++++ src/security/security_util.h | 32 ++++ src/util/virfile.c | 121 ++++++++++++++ src/util/virfile.h | 11 ++ tools/Makefile.am | 1 + tools/libvirt_recover_xattrs.sh | 89 ++++++++++ 21 files changed, 857 insertions(+), 150 deletions(-) create mode 100644 src/security/security_util.c create mode 100644 src/security/security_util.h create mode 100755 tools/libvirt_recover_xattrs.sh -- 2.18.1

2 40

[libvirt] [PATCH 0/2] Fix 2 small LXC network interface bugs
by Laine Stump 06 Dec '18

06 Dec '18

(Patch 1 fixes a recently filed BZ that it just happened was easy for me to test, because I had encountered the same problem awhile back) Laine Stump (2): lxc: stop incorrectly validating interface type lxc: check actual type of interface not config type src/lxc/lxc_controller.c | 20 +++++++++++++------- 1 file changed, 13 insertions(+), 7 deletions(-) -- 2.19.2

2 6

Re: [libvirt] [Qemu-devel] [RFC 0/2] Attempt to implement the standby feature for assigned network devices
by Michael Roth 05 Dec '18

05 Dec '18

Quoting Sameeh Jubran (2018-10-25 13:01:10) > On Thu, Oct 25, 2018 at 5:06 PM Sameeh Jubran <sameeh(a)daynix.com> wrote: > > > > From: Sameeh Jubran <sjubran(a)redhat.com> > > > > Hi all, > > > > Background: > > > > There has been a few attempts to implement the standby feature for vfio > > assigned devices which aims to enable the migration of such devices. This > > is another attempt. > > > > The series implements an infrastructure for hiding devices from the bus > > upon boot. What it does is the following: > > > > * In the first patch the infrastructure for hiding the device is added > > for the qbus and qdev APIs. A "hidden" boolean is added to the device > > state and it is set based on a callback to the standby device which > > registers itself for handling the assessment: "should the primary device > > be hidden?" by cross validating the ids of the devices. > > > > * In the second patch the virtio-net uses the API to hide the vfio > > device and unhides it when the feature is acked. > > > > Disclaimers: > > > > * I have only scratch tested this and from qemu side, it seems to be > > working. > > * This is an RFC so it lacks some proper error handling in few cases > > and proper resource freeing. I wanted to get some feedback first > > before it is finalized. > > > > Command line example: > > > > /home/sameeh/Builds/failover/qemu/x86_64-softmmu/qemu-system-x86_64 \ > > -netdev tap,id=hostnet0,script=world_bridge_standalone.sh,downscript=no,ifname=cc1_71 \ > > -netdev tap,vhost=on,id=hostnet1,script=world_bridge_standalone.sh,downscript=no,ifname=cc1_72,queues=4 \ > > -device virtio-net,host_mtu=1500,netdev=hostnet1,id=cc1_72,vectors=10,mq=on,primary=cc1_71 \ > > -device e1000,netdev=hostnet0,id=cc1_71,standby=cc1_72 \ > > > > Migration support: > > > > Pre migration or during setup phase of the migration we should send an > > unplug request to the guest to unplug the primary device. I haven't had > > the chance to implement that part yet but should do soon. Do you know > > what's the best approach to do so? I wanted to have a callback to the > > virtio-net device which tries to send an unplug request to the guest and > > if succeeds then the migration continues. It needs to handle the case where > > the migration fails and then it has to replug the primary device back. > I think that the "add_migration_state_change_notifier" API call can be used > from within the virtio-net device to achieve this, what do you think? I think it would be good to hear from the libvirt folks (on Cc:) on this as having QEMU unplug a device without libvirt's involvement seems like it could cause issues. Personally I think it seems cleaner to just have QEMU handle the 'hidden' aspects of the device and leave it to QMP/libvirt to do the unplug beforehand. On the libvirt side I could imagine adding an option like virsh migrate --switch-to-standby-networking or something along that line to do it automatically (if we decide doing it automatically is even needed on that end). > > > > The following terms are used as interchangeable: > > standby - virtio-net > > primary - vfio-device - physical device - assigned device > > > > Please share your thoughts and suggestions, > > Thanks! > > > > Sameeh Jubran (2): > > qdev/qbus: Add hidden device support > > virtio-net: Implement VIRTIO_NET_F_STANDBY feature > > > > hw/core/qdev.c | 48 +++++++++++++++++++++++++--- > > hw/net/virtio-net.c | 25 +++++++++++++++ > > hw/pci/pci.c | 1 + > > include/hw/pci/pci.h | 2 ++ > > include/hw/qdev-core.h | 11 ++++++- > > include/hw/virtio/virtio-net.h | 5 +++ > > qdev-monitor.c | 58 ++++++++++++++++++++++++++++++++-- > > 7 files changed, 142 insertions(+), 8 deletions(-) > > > > -- > > 2.17.0 > > > > > -- > Respectfully, > Sameeh Jubran > Linkedin > Software Engineer @ Daynix. >

4 4

[libvirt] [PATCH] qemu: Add support for postcopy-requests migration statistics
by Jiri Denemark 05 Dec '18

05 Dec '18

QEMU can report how many times during post-copy migration the domain running on the destination host tried to access a page which has not been migrated yet. Signed-off-by: Jiri Denemark <jdenemar(a)redhat.com> --- include/libvirt/libvirt-domain.h | 10 ++++++++++ src/qemu/qemu_domain.c | 5 ++++- src/qemu/qemu_migration_cookie.c | 5 +++++ src/qemu/qemu_monitor.h | 1 + src/qemu/qemu_monitor_json.c | 2 ++ tools/virsh-domain.c | 8 ++++++++ 6 files changed, 30 insertions(+), 1 deletion(-) diff --git a/include/libvirt/libvirt-domain.h b/include/libvirt/libvirt-domain.h index 71debd92b8..897b06e6d0 100644 --- a/include/libvirt/libvirt-domain.h +++ b/include/libvirt/libvirt-domain.h @@ -3378,6 +3378,16 @@ typedef enum { */ # define VIR_DOMAIN_JOB_MEMORY_ITERATION "memory_iteration" +/** + * VIR_DOMAIN_JOB_MEMORY_POSTCOPY_REQS: + * + * virDomainGetJobStats field: number page requests received from the + * destination host during post-copy migration, as VIR_TYPED_PARAM_ULLONG. + * This counter is incremented whenever the migrated domain tries to access + * a memory page which has not been transferred from the source host yet. + */ +# define VIR_DOMAIN_JOB_MEMORY_POSTCOPY_REQS "memory_postcopy_requests" + /** * VIR_DOMAIN_JOB_DISK_TOTAL: * diff --git a/src/qemu/qemu_domain.c b/src/qemu/qemu_domain.c index fbe63e2e1d..3dbd9d47ca 100644 --- a/src/qemu/qemu_domain.c +++ b/src/qemu/qemu_domain.c @@ -609,7 +609,10 @@ qemuDomainMigrationJobInfoToParams(qemuDomainJobInfoPtr jobInfo, stats->ram_dirty_rate) < 0 || virTypedParamsAddULLong(&par, &npar, &maxpar, VIR_DOMAIN_JOB_MEMORY_ITERATION, - stats->ram_iteration) < 0) + stats->ram_iteration) < 0 || + virTypedParamsAddULLong(&par, &npar, &maxpar, + VIR_DOMAIN_JOB_MEMORY_POSTCOPY_REQS, + stats->ram_postcopy_reqs) < 0) goto error; if (stats->ram_page_size > 0 && diff --git a/src/qemu/qemu_migration_cookie.c b/src/qemu/qemu_migration_cookie.c index 60df449d53..84f0101382 100644 --- a/src/qemu/qemu_migration_cookie.c +++ b/src/qemu/qemu_migration_cookie.c @@ -703,6 +703,9 @@ qemuMigrationCookieStatisticsXMLFormat(virBufferPtr buf, virBufferAsprintf(buf, "<%1$s>%2$llu</%1$s>\n", VIR_DOMAIN_JOB_MEMORY_ITERATION, stats->ram_iteration); + virBufferAsprintf(buf, "<%1$s>%2$llu</%1$s>\n", + VIR_DOMAIN_JOB_MEMORY_POSTCOPY_REQS, + stats->ram_postcopy_reqs); virBufferAsprintf(buf, "<%1$s>%2$llu</%1$s>\n", VIR_DOMAIN_JOB_MEMORY_PAGE_SIZE, @@ -1102,6 +1105,8 @@ qemuMigrationCookieStatisticsXMLParse(xmlXPathContextPtr ctxt) ctxt, &stats->ram_dirty_rate); virXPathULongLong("string(./" VIR_DOMAIN_JOB_MEMORY_ITERATION "[1])", ctxt, &stats->ram_iteration); + virXPathULongLong("string(./" VIR_DOMAIN_JOB_MEMORY_POSTCOPY_REQS "[1])", + ctxt, &stats->ram_postcopy_reqs); virXPathULongLong("string(./" VIR_DOMAIN_JOB_MEMORY_PAGE_SIZE "[1])", ctxt, &stats->ram_page_size); diff --git a/src/qemu/qemu_monitor.h b/src/qemu/qemu_monitor.h index 48b142a4f4..3430426c4b 100644 --- a/src/qemu/qemu_monitor.h +++ b/src/qemu/qemu_monitor.h @@ -706,6 +706,7 @@ struct _qemuMonitorMigrationStats { unsigned long long ram_dirty_rate; unsigned long long ram_page_size; unsigned long long ram_iteration; + unsigned long long ram_postcopy_reqs; unsigned long long disk_transferred; unsigned long long disk_remaining; diff --git a/src/qemu/qemu_monitor_json.c b/src/qemu/qemu_monitor_json.c index 3de298c9e2..4c151dd405 100644 --- a/src/qemu/qemu_monitor_json.c +++ b/src/qemu/qemu_monitor_json.c @@ -3296,6 +3296,8 @@ qemuMonitorJSONGetMigrationStatsReply(virJSONValuePtr reply, &stats->ram_page_size)); ignore_value(virJSONValueObjectGetNumberUlong(ram, "dirty-sync-count", &stats->ram_iteration)); + ignore_value(virJSONValueObjectGetNumberUlong(ram, "postcopy-requests", + &stats->ram_postcopy_reqs)); disk = virJSONValueObjectGetObject(ret, "disk"); if (disk) { diff --git a/tools/virsh-domain.c b/tools/virsh-domain.c index 372bdb95d3..a71cc9d0be 100644 --- a/tools/virsh-domain.c +++ b/tools/virsh-domain.c @@ -6233,6 +6233,14 @@ cmdDomjobinfo(vshControl *ctl, const vshCmd *cmd) } else if (rc) { vshPrint(ctl, "%-17s %-12llu\n", _("Iteration:"), value); } + + if ((rc = virTypedParamsGetULLong(params, nparams, + VIR_DOMAIN_JOB_MEMORY_POSTCOPY_REQS, + &value)) < 0) { + goto save_error; + } else if (rc) { + vshPrint(ctl, "%-17s %-12llu\n", _("Postcopy requests:"), value); + } } if (info.fileTotal || info.fileRemaining || info.fileProcessed) { -- 2.19.1

2 1

[libvirt] [PATCH 0/5] Various cleanups and fixes (blockdev-add saga)
by Peter Krempa 05 Dec '18

05 Dec '18

Few random patches extracted from my blockdev series. Peter Krempa (5): tests: qemuxml2xml: Add few debug statements for status XML testing util: xml: Always consume args of virXMLFormatElement qemu: domain: Initialize proper element in qemuDomainPrepareStorageSourceBlockdev conf: snapshot: Remove file format check from parser tests: qemuxml2argv: Remove disks from few tests which don't need them src/conf/domain_conf.c | 12 ++---------- src/conf/snapshot_conf.c | 12 +++--------- src/qemu/qemu_domain.c | 9 +++------ src/util/virxml.c | 16 ++++++++++++---- .../channel-unix-guestfwd.x86_64-2.5.0.args | 2 -- .../channel-unix-guestfwd.x86_64-latest.args | 2 -- tests/qemuxml2argvdata/channel-unix-guestfwd.xml | 5 ----- .../console-virtio-unix.x86_64-2.5.0.args | 2 -- .../console-virtio-unix.x86_64-latest.args | 2 -- tests/qemuxml2argvdata/console-virtio-unix.xml | 5 ----- tests/qemuxml2argvdata/net-vhostuser.args | 3 --- .../net-vhostuser.x86_64-2.5.0.args | 2 -- .../net-vhostuser.x86_64-latest.args | 2 -- tests/qemuxml2argvdata/net-vhostuser.xml | 6 ------ .../parallel-unix-chardev.x86_64-2.5.0.args | 2 -- .../parallel-unix-chardev.x86_64-latest.args | 2 -- tests/qemuxml2argvdata/parallel-unix-chardev.xml | 5 ----- tests/qemuxml2xmloutdata/net-vhostuser.xml | 6 ------ tests/qemuxml2xmltest.c | 9 ++++++--- 19 files changed, 26 insertions(+), 78 deletions(-) -- 2.19.2

2 6

[libvirt] [PATCH v2] qemuDomainRemoveRNGDevice: Remove associated chardev too
by Michal Privoznik 05 Dec '18

05 Dec '18

https://bugzilla.redhat.com/show_bug.cgi?id=1656014 An RNG device can consists of more devices than RND device itself. For instance, in case of EGD there is a chardev that connects to EGD daemon and feeds the qemu with random data. When doing RNG device removal we have to remove the associated chardev as well. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- diff to v1: - Detach Extension Device too - Reoder the way in which devices are detached (as suggested in v1) src/qemu/qemu_hotplug.c | 16 +++++++++++++--- 1 file changed, 13 insertions(+), 3 deletions(-) diff --git a/src/qemu/qemu_hotplug.c b/src/qemu/qemu_hotplug.c index 5f756b7267..f5f8cec59e 100644 --- a/src/qemu/qemu_hotplug.c +++ b/src/qemu/qemu_hotplug.c @@ -4814,7 +4814,7 @@ qemuDomainRemoveRNGDevice(virQEMUDriverPtr driver, qemuDomainObjPrivatePtr priv = vm->privateData; ssize_t idx; int ret = -1; - int rc; + int rc = 0; VIR_DEBUG("Removing RNG device %s from domain %p %s", rng->info.alias, vm, vm->def->name); @@ -4828,7 +4828,17 @@ qemuDomainRemoveRNGDevice(virQEMUDriverPtr driver, qemuDomainObjEnterMonitor(driver, vm); - rc = qemuMonitorDelObject(priv->mon, objAlias); + if (qemuDomainDetachExtensionDevice(priv->mon, &rng->info) < 0) + rc = -1; + + if (rc == 0 && + qemuMonitorDelObject(priv->mon, objAlias) < 0) + rc = -1; + + if (rng->backend == VIR_DOMAIN_RNG_BACKEND_EGD && + rc == 0 && + qemuMonitorDetachCharDev(priv->mon, charAlias) < 0) + rc = -1; if (qemuDomainObjExitMonitor(driver, vm) < 0) goto cleanup; @@ -4837,7 +4847,7 @@ qemuDomainRemoveRNGDevice(virQEMUDriverPtr driver, rc == 0 && qemuDomainDelChardevTLSObjects(driver, vm, rng->source.chardev, charAlias) < 0) - goto cleanup; + rc = -1; virDomainAuditRNG(vm, rng, NULL, "detach", rc == 0); -- 2.19.2

2 1

[libvirt] [PATCH libvirt-python] virNodeInfo.memory is in KiB
by Philipp Hahn 05 Dec '18

05 Dec '18

but the Python library does an extra left shift of 10 bits returning MiB instead: > # cat y.c > #include <stdlib.h> > #include <stdio.h> > #include <libvirt.h> > int main(void) { > virConnectPtr conn = virConnectOpen("qemu:///system"); > virNodeInfo info; > int rv = virNodeGetInfo(conn, &info); > printf("%ld\n", info.memory); > return rv; > } > # gcc y.c -I/usr/include/libvirt -lvirt > # ./a.out > 4041088 > # python -c 'import libvirt;c=libvirt.open("qemu:///system");print(c.getInfo()[1])' > 3946 Fixes: 197153c6 Signed-off-by: Philipp Hahn <hahn(a)univention.de> --- libvirt-override.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/libvirt-override.c b/libvirt-override.c index f7b2f6b..616fa1c 100644 --- a/libvirt-override.c +++ b/libvirt-override.c @@ -2740,7 +2740,7 @@ libvirt_virNodeGetInfo(PyObject *self ATTRIBUTE_UNUSED, VIR_PY_LIST_SET_GOTO(py_retval, 0, libvirt_constcharPtrWrap(&info.model[0]), error); VIR_PY_LIST_SET_GOTO(py_retval, 1, - libvirt_longWrap((long) info.memory >> 10), error); + libvirt_longWrap((long) info.memory), error); VIR_PY_LIST_SET_GOTO(py_retval, 2, libvirt_intWrap((int) info.cpus), error); VIR_PY_LIST_SET_GOTO(py_retval, 3, libvirt_intWrap((int) info.mhz), error); VIR_PY_LIST_SET_GOTO(py_retval, 4, libvirt_intWrap((int) info.nodes), error); -- 2.11.0

2 2

[libvirt] [PATCH RFC v2] qemu: fix deadlock when waiting in non async jobs
by Nikolay Shirokovskiy 05 Dec '18

05 Dec '18

Block job abort operation can not handle properly qemu crashes when waiting for abort/pivot completion. Deadlock scenario is next: - qemuDomainBlockJobAbort waits for pivot/abort completion - qemu crashes, then qemuProcessBeginStopJob broadcasts for VM condition and then waits for job condition (taken by qemuDomainBlockJobAbort) - qemuDomainBlockJobAbort awakes but nothing really changed, VM is still active (vm->def->id != -1) so thread starts waiting for completion again. Now two threads are in deadlock. First let's remove broadcast in qemuProcessBeginStopJob. It is simply wrong because it is not set any condition before broadcast so that awaked threads can not detect any changes. Crashing domain during async job will continue to be handled properly because destroy job can run concurrently with async job and destroy job calls qemuProcessStop which sets vm->def->id to -1 and broadcasts. Second let's introduce flag that EOF is received and broadcast after that. Now non async jobs can check this flag in wait loop. Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy(a)virtuozzo.com> --- Diff from v1: - patches 1 and 2 are already merged - don't bother with reporting monitor EOF reason to user as most of time it is simply "unexpected eof" (this implies dropping patch 3) - drop patch 5 as we now always report "domain is being stopped" in qemuDomainObjWait - don't signal on monitor error for simplicity (otherwise we need to report something more elaborate that "domain is being stopped" as we don't kill domain on monitor errors. On the other hand I guess monitor error is rare case to handle it right now) - keep virDomainObjWait for async jobs It's a bit uneven that for async jobs domain is destroyed concurrently and for non async jobs it will be actually destroyed after job get completed. Also if non async job needs issuing commands to qemu on cleanup then we will send these commands in vain polluting logs etc because qemu process in not running at this moment but typical check (virDomainObjIsActive) will think it is still running. Domain is destroyed (qemuProcessStop) in a job due to patches [1] and [2]. However AFAIU it is not neccessary. If qemuProcessStop does not drop VM lock then we don't need extra job to make qemuProcessStop and main job not interleave. And we can drop the lock now only in qemuDomainObjBeginNestedJob in qemuProcessStop which is introduced in [2]. AFAIU we can fix issues mentioned in [2] the other way for example like it is done for qemu agent - we save agent monitor reference on stack for entering/exiting agent monitor. So I wonder can we instead of this fix remove job for qemuProcessStop and run destroying domain cuncurrently for non async jobs too. [1] commit 8c9ff9960b29d4703a99efdd1cadcf6f48799cc0 Author: Jiri Denemark <jdenemar(a)redhat.com> Date: Thu Feb 11 15:32:48 2016 +0100 qemu: Process monitor EOF in a job [2] commit 81f50cb92d16643bcd749e3ab5b404b8b7cec643 Author: Jiri Denemark <jdenemar(a)redhat.com> Date: Thu Feb 11 11:20:28 2016 +0100 qemu: Avoid calling qemuProcessStop without a job src/qemu/qemu_domain.c | 39 +++++++++++++++++++++++++++++++++++++++ src/qemu/qemu_domain.h | 4 ++++ src/qemu/qemu_driver.c | 2 +- src/qemu/qemu_hotplug.c | 4 ++-- src/qemu/qemu_process.c | 9 +++++---- 5 files changed, 51 insertions(+), 7 deletions(-) diff --git a/src/qemu/qemu_domain.c b/src/qemu/qemu_domain.c index 939b2a3..aead72b 100644 --- a/src/qemu/qemu_domain.c +++ b/src/qemu/qemu_domain.c @@ -13534,3 +13534,42 @@ qemuDomainRunningReasonToResumeEvent(virDomainRunningReason reason) return VIR_DOMAIN_EVENT_RESUMED_UNPAUSED; } + + +/** + * Waits for domain condition to be triggered for a specific period of time. + * if @until is 0 then waits indefinetely. + * + * Returns: + * -1 on error + * 0 on success + * 1 on timeout + */ +int +qemuDomainObjWait(virDomainObjPtr vm, unsigned long long until) +{ + qemuDomainObjPrivatePtr priv = vm->privateData; + int rc; + + if (until) + rc = virCondWaitUntil(&vm->cond, &vm->parent.lock, until); + else + rc = virCondWait(&vm->cond, &vm->parent.lock); + + if (rc < 0) { + if (until && errno == ETIMEDOUT) + return 1; + + virReportSystemError(errno, "%s", + _("failed to wait for domain condition")); + return -1; + } + + if (priv->monEOF) { + virReportError(VIR_ERR_OPERATION_FAILED, "%s", + _("domain is being stopped")); + return -1; + } + + return 0; +} diff --git a/src/qemu/qemu_domain.h b/src/qemu/qemu_domain.h index 2f8a1bf..36ab294 100644 --- a/src/qemu/qemu_domain.h +++ b/src/qemu/qemu_domain.h @@ -281,6 +281,7 @@ struct _qemuDomainObjPrivate { virDomainChrSourceDefPtr monConfig; bool monJSON; bool monError; + bool monEOF; unsigned long long monStart; qemuAgentPtr agent; @@ -1085,4 +1086,7 @@ void qemuDomainStorageIdReset(qemuDomainObjPrivatePtr priv); virDomainEventResumedDetailType qemuDomainRunningReasonToResumeEvent(virDomainRunningReason reason); +int +qemuDomainObjWait(virDomainObjPtr vm, unsigned long long until); + #endif /* __QEMU_DOMAIN_H__ */ diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index b238309..f4250da 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -17142,7 +17142,7 @@ qemuDomainBlockJobAbort(virDomainPtr dom, qemuDomainDiskPrivatePtr diskPriv = QEMU_DOMAIN_DISK_PRIVATE(disk); qemuBlockJobUpdate(vm, QEMU_ASYNC_JOB_NONE, disk, NULL); while (diskPriv->blockjob) { - if (virDomainObjWait(vm) < 0) { + if (qemuDomainObjWait(vm, 0) < 0) { ret = -1; goto endjob; } diff --git a/src/qemu/qemu_hotplug.c b/src/qemu/qemu_hotplug.c index 4558a3c..8189629 100644 --- a/src/qemu/qemu_hotplug.c +++ b/src/qemu/qemu_hotplug.c @@ -165,7 +165,7 @@ qemuHotplugWaitForTrayEject(virDomainObjPtr vm, return -1; while (disk->tray_status != VIR_DOMAIN_DISK_TRAY_OPEN) { - if ((rc = virDomainObjWaitUntil(vm, now + CHANGE_MEDIA_TIMEOUT)) < 0) + if ((rc = qemuDomainObjWait(vm, now + CHANGE_MEDIA_TIMEOUT)) < 0) return -1; if (rc > 0) { @@ -5002,7 +5002,7 @@ qemuDomainWaitForDeviceRemoval(virDomainObjPtr vm) until += qemuDomainRemoveDeviceWaitTime; while (priv->unplug.alias) { - if ((rc = virDomainObjWaitUntil(vm, until)) == 1) + if ((rc = qemuDomainObjWait(vm, until)) == 1) return 0; if (rc < 0) { diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index 29b0ba1..dd03269 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -290,9 +290,12 @@ qemuProcessHandleMonitorEOF(qemuMonitorPtr mon, virObjectLock(vm); + priv = vm->privateData; + priv->monEOF = true; + virDomainObjBroadcast(vm); + VIR_DEBUG("Received EOF on %p '%s'", vm, vm->def->name); - priv = vm->privateData; if (priv->beingDestroyed) { VIR_DEBUG("Domain is being destroyed, EOF is expected"); goto cleanup; @@ -5996,6 +5999,7 @@ qemuProcessPrepareDomain(virQEMUDriverPtr driver, priv->monJSON = true; priv->monError = false; + priv->monEOF = false; priv->monStart = 0; priv->gotShutdown = false; priv->runningReason = VIR_DOMAIN_RUNNING_UNKNOWN; @@ -6965,9 +6969,6 @@ qemuProcessBeginStopJob(virQEMUDriverPtr driver, if (qemuProcessKill(vm, killFlags) < 0) goto cleanup; - /* Wake up anything waiting on domain condition */ - virDomainObjBroadcast(vm); - if (qemuDomainObjBeginJob(driver, vm, job) < 0) goto cleanup; -- 1.8.3.1

2 5