- Devel - libvirt List Archives

[libvirt] [PATCH] docs: formatdomain: Note the caveats for CPU policy option "force"
by Kashyap Chamarthy 22 Jun '18

22 Jun '18

Eduardo Habkost has pointed out that the current documentation of libvirt's CPU feature policy "require" vs. "force" does not match QEMU's behaviour. Update the documentation by spelling out the QEMU version dependency and explain in which scenarios the usage of "policy = 'force'" is applicable or not. Signed-off-by: Kashyap Chamarthy <kchamart(a)redhat.com> --- Wordsmithing / corrections welcome. --- docs/formatdomain.html.in | 13 +++++++++++-- 1 file changed, 11 insertions(+), 2 deletions(-) diff --git a/docs/formatdomain.html.in b/docs/formatdomain.html.in index 6912762f28..4d6c3892ee 100644 --- a/docs/formatdomain.html.in +++ b/docs/formatdomain.html.in @@ -1566,8 +1566,17 @@ <dl> <dt><code>force</code></dt> - <dd>The virtual CPU will claim the feature is supported regardless - of it being supported by host CPU.</dd> + <dd>The virtual CPU will claim the feature is supported + regardless of it being supported by host CPU -- this is only + true for QEMU version older than 2.9.0. I.e. when using the + CPU mode 'host-model', libvirt identifies which CPU features + to use by looking at host CPUID. For that to take effect, it + is mandatory to use <code>force</code> to tell libvirt that a + said CPU feature must be used despite it not existing in the + host -- this applicable only for a very limited set of CPU + features, such as 'x2apic', virt-ssbd' (for AMD CPUs).</dd> + <dd>However, when using QEMU 2.9.0 and above, there should + never be any need to use <code>force</code>.</dd> <dt><code>require</code></dt> <dd>Guest creation will fail unless the feature is supported by the host CPU or the hypervisor is able to emulate it.</dd> -- 2.17.0

3 3

[libvirt] [PATCH v3 00/20] nwfilter: refactor the driver to make it independent of virt drivers
by Daniel P. Berrangé 22 Jun '18

22 Jun '18

v1: https://www.redhat.com/archives/libvir-list/2018-April/msg02616.html v2: https://www.redhat.com/archives/libvir-list/2018-May/msg01145.html Today the nwfilter driver is entangled with the virt drivers in both directions. At various times when rebuilding filters nwfilter will call out to the virt driver to iterate over running guest's NICs. This has caused very complicated lock ordering rules to be required. If we are to split the virt drivers out into separate daemons we need to get rid of this coupling since we don't want the separate daemons calling each other, as that risks deadlock if all of the RPC workers are busy. The obvious way to solve this is to have the nwfilter driver remember all the filters it has active, avoiding the need to iterate over running guests. Still todo - Document the new XML format Changed in v3: - Updated API version numbers - Use accessors for virNWFilterBindingObjPtr struct - Other fixes John mentioned Changed in v2: - The virNWFilterBindingPtr was renamed virNWFilterBindingDefPtr - New virNWFilterBindingObjPtr & virNWFilterBindingObjListPtr structs added to track the objects in the driver - New virNWFilterBindingPtr public API type was added - New public APIs for listing filter bindings, querying XML, and creating/deleting them - Convert the virt drivers to use the public API for creating and deleting bindings - Persistent active bindings out to disk so they're preserved across restarts - Added RNG schema and XML-2-XML test - New virsh commands for listing/querying XML/creating/deleting bindings Daniel P. Berrangé (20): conf: change virNWFilterBindingPtr to virNWFilterBindingDefPtr conf: add missing virxml.h include for nwfilter_params.h conf: move virNWFilterBindingDefPtr into its own files conf: add support for parsing/formatting virNWFilterBindingDefPtr schemas: add schema for nwfilter binding XML document nwfilter: export port binding concept in the public API access: add nwfilter binding object permissions remote: add support for nwfilter binding objects virsh: add nwfilter binding commands nwfilter: convert the gentech driver code to use virNWFilterBindingDefPtr nwfilter: convert IP address learning code to virNWFilterBindingDefPtr nwfilter: convert DHCP address snooping code to virNWFilterBindingDefPtr conf: report an error if nic needs filtering by no driver is present conf: introduce a virNWFilterBindingObjPtr struct conf: introduce a virNWFilterBindingObjListPtr struct nwfilter: keep track of active filter bindings nwfilter: remove virt driver callback layer for rebuilding filters nwfilter: wire up new APIs for listing and querying filter bindings nwfilter: wire up new APIs for creating and deleting nwfilter bindings nwfilter: convert virt drivers to use public API for nwfilter bindings docs/schemas/domaincommon.rng | 27 +- docs/schemas/nwfilter.rng | 29 +- docs/schemas/nwfilter_params.rng | 32 ++ docs/schemas/nwfilterbinding.rng | 49 ++ include/libvirt/libvirt-nwfilter.h | 39 ++ include/libvirt/virterror.h | 2 + src/access/viraccessdriver.h | 5 + src/access/viraccessdrivernop.c | 10 + src/access/viraccessdriverpolkit.c | 21 + src/access/viraccessdriverstack.c | 24 + src/access/viraccessmanager.c | 15 + src/access/viraccessmanager.h | 5 + src/access/viraccessperm.c | 7 +- src/access/viraccessperm.h | 38 ++ src/conf/Makefile.inc.am | 6 + src/conf/domain_nwfilter.c | 125 ++++- src/conf/domain_nwfilter.h | 13 - src/conf/nwfilter_conf.c | 188 +------ src/conf/nwfilter_conf.h | 68 +-- src/conf/nwfilter_params.h | 1 + src/conf/virnwfilterbindingdef.c | 280 ++++++++++ src/conf/virnwfilterbindingdef.h | 65 +++ src/conf/virnwfilterbindingobj.c | 299 +++++++++++ src/conf/virnwfilterbindingobj.h | 69 +++ src/conf/virnwfilterbindingobjlist.c | 487 ++++++++++++++++++ src/conf/virnwfilterbindingobjlist.h | 69 +++ src/conf/virnwfilterobj.c | 4 +- src/conf/virnwfilterobj.h | 4 + src/datatypes.c | 67 +++ src/datatypes.h | 31 ++ src/driver-nwfilter.h | 30 ++ src/libvirt-nwfilter.c | 305 +++++++++++ src/libvirt_private.syms | 45 +- src/libvirt_public.syms | 9 + src/lxc/lxc_driver.c | 28 - src/nwfilter/nwfilter_dhcpsnoop.c | 158 +++--- src/nwfilter/nwfilter_dhcpsnoop.h | 7 +- src/nwfilter/nwfilter_driver.c | 218 ++++++-- src/nwfilter/nwfilter_gentech_driver.c | 337 ++++++------ src/nwfilter/nwfilter_gentech_driver.h | 22 +- src/nwfilter/nwfilter_learnipaddr.c | 104 ++-- src/nwfilter/nwfilter_learnipaddr.h | 7 +- src/qemu/qemu_driver.c | 25 - src/remote/remote_daemon_dispatch.c | 15 + src/remote/remote_driver.c | 20 + src/remote/remote_protocol.x | 90 +++- src/remote_protocol-structs | 43 ++ src/rpc/gendispatch.pl | 15 +- src/uml/uml_driver.c | 29 -- src/util/virerror.c | 12 + tests/Makefile.am | 7 + .../filter-vars.xml | 11 + .../virnwfilterbindingxml2xmldata/simple.xml | 9 + tests/virnwfilterbindingxml2xmltest.c | 112 ++++ tests/virschematest.c | 1 + tools/virsh-completer.c | 45 ++ tools/virsh-completer.h | 4 + tools/virsh-nwfilter.c | 317 ++++++++++++ tools/virsh-nwfilter.h | 8 + 59 files changed, 3283 insertions(+), 829 deletions(-) create mode 100644 docs/schemas/nwfilter_params.rng create mode 100644 docs/schemas/nwfilterbinding.rng create mode 100644 src/conf/virnwfilterbindingdef.c create mode 100644 src/conf/virnwfilterbindingdef.h create mode 100644 src/conf/virnwfilterbindingobj.c create mode 100644 src/conf/virnwfilterbindingobj.h create mode 100644 src/conf/virnwfilterbindingobjlist.c create mode 100644 src/conf/virnwfilterbindingobjlist.h create mode 100644 tests/virnwfilterbindingxml2xmldata/filter-vars.xml create mode 100644 tests/virnwfilterbindingxml2xmldata/simple.xml create mode 100644 tests/virnwfilterbindingxml2xmltest.c -- 2.17.0

2 37

[libvirt] [PATCH] qemu: fix msg could be a wild pointer in qemuMonitorIOProcess()
by Weilun Zhu 22 Jun '18

22 Jun '18

As qemuMonitorJSONIOProcess will call qemuMonitorJSONIOProcessEvent which unlocks the monitor mutex, there is some extreme situation, eg qemu send message to monitor twice in a short time, where the local viriable 'msg' of qemuMonitorIOProcess could be a wild point: 1. qemuMonitorSend() assign mon->msg to parameter 'msg', which is alse a local variable of its caller qemuMonitorJSONCommandWithFd(), cause eventloop to send message to monitor, then wait condition. 2. qemu send message to monitor for the first time immediately. 3. qemuMonitorIOProcess() is called, then wake up the qemuMonitorSend() thread, but the qemuMonitorSend() thread stuck for a while as cpu pressure or some other reasons, which means the qemu monitor is still unlocked. 4. qemu send event message to monitor for the second time, such as RTC_CHANGE event 5. qemuMonitorIOProcess() is called again, the local viriable 'msg' is assigned to mon->msg. 6. qemuMonitorIOProcess() call qemuMonitorJSONIOProcess() to deal with the qemu event. 7. qemuMonitorJSONIOProcess() unlock the qemu monitor in the macro 'QEMU_MONITOR_CALLBACK', then qemuMonitorSend() thread get the mutex and free the mon->msg, assign mon->msg to NULL. Signed-off-by: Weilun Zhu <zhuweilun(a)huawei.com> --- src/qemu/qemu_monitor.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/src/qemu/qemu_monitor.c b/src/qemu/qemu_monitor.c index d6771c1d52..6ed475ede0 100644 --- a/src/qemu/qemu_monitor.c +++ b/src/qemu/qemu_monitor.c @@ -466,7 +466,11 @@ qemuMonitorIOProcess(qemuMonitorPtr mon) #if DEBUG_IO VIR_DEBUG("Process done %d used %d", (int)mon->bufferOffset, len); #endif - if (msg && msg->finished) + + /* As the monitor mutex was unlocked in qemuMonitorJSONIOProcess() + * while dealing with qemu event, mon->msg could be changed which + * means the above 'msg' may be invalid, thus we use 'mon->msg' here */ + if (mon->msg && mon->msg->finished) virCondBroadcast(&mon->notify); return len; } -- 2.18.0.rc1

2 1

[libvirt] [PATCH] news: Document recent agent job change
by Michal Privoznik 22 Jun '18

22 Jun '18

Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- docs/news.xml | 13 +++++++++++++ 1 file changed, 13 insertions(+) diff --git a/docs/news.xml b/docs/news.xml index 08e5dcbda3..a3eea4eff6 100644 --- a/docs/news.xml +++ b/docs/news.xml @@ -71,6 +71,19 @@ Capabilities XML now provide information about host IOMMU support. </description> </change> + <change> + <summary> + qemu: Allow concurrent access to monitor and guest agent + </summary> + <description> + Historically libvirt prevented concurrent accesses to + the qemu monitor and the guest agent. Therefore two + independent calls (one querying the monitor and the + other querying guest agent) would serialize which hurts + performance. The code was reworked to allow two + independent calls run at the same time. + </description> + </change> </section> <section title="Bug fixes"> <change> -- 2.16.4

2 1

[libvirt] [PATCH] vmx: allow an odd number of vCPUs
by Pino Toscano 22 Jun '18

22 Jun '18

Most probably this was a limitation in older ESX versions, and it seems it does not exist anymore in more recent versions; see the following thread: https://www.redhat.com/archives/libvir-list/2018-May/msg02159.html https://www.redhat.com/archives/libvir-list/2018-June/msg00043.html Hence, allow an odd number (greater than 1) of vCPUs, since most probably older versions of ESXi will error out anyway. https://bugzilla.redhat.com/show_bug.cgi?id=1584091 Signed-off-by: Pino Toscano <ptoscano(a)redhat.com> --- src/vmx/vmx.c | 15 +++++++-------- 1 file changed, 7 insertions(+), 8 deletions(-) diff --git a/src/vmx/vmx.c b/src/vmx/vmx.c index 35b4807d22..fe24b060d7 100644 --- a/src/vmx/vmx.c +++ b/src/vmx/vmx.c @@ -59,7 +59,7 @@ def->name = <value> <=> displayName = "<value>" def->mem.max_balloon = <value kilobyte> <=> memsize = "<value megabyte>" # must be a multiple of 4, defaults to 32 def->mem.cur_balloon = <value kilobyte> <=> sched.mem.max = "<value megabyte>" # defaults to "unlimited" -> def->mem.cur_balloon = def->mem.max_balloon def->mem.min_guarantee = <value kilobyte> <=> sched.mem.minsize = "<value megabyte>" # defaults to 0 -def->maxvcpus = <value> <=> numvcpus = "<value>" # must be 1 or a multiple of 2, defaults to 1 +def->maxvcpus = <value> <=> numvcpus = "<value>" # must be greater than 0, defaults to 1 def->cpumask = <uint list> <=> sched.cpu.affinity = "<uint list>" def->cputune.shares = <value> <=> sched.cpu.shares = "<value>" # with handling for special values # "high", "normal", "low" @@ -1452,10 +1452,10 @@ virVMXParseConfig(virVMXContext *ctx, if (virVMXGetConfigLong(conf, "numvcpus", &numvcpus, 1, true) < 0) goto cleanup; - if (numvcpus <= 0 || (numvcpus % 2 != 0 && numvcpus != 1)) { + if (numvcpus <= 0) { virReportError(VIR_ERR_INTERNAL_ERROR, _("Expecting VMX entry 'numvcpus' to be an unsigned " - "integer (1 or a multiple of 2) but found %lld"), numvcpus); + "integer greater than 0 but found %lld"), numvcpus); goto cleanup; } @@ -3183,11 +3183,10 @@ virVMXFormatConfig(virVMXContext *ctx, virDomainXMLOptionPtr xmlopt, virDomainDe goto cleanup; } maxvcpus = virDomainDefGetVcpusMax(def); - if (maxvcpus == 0 || (maxvcpus % 2 != 0 && maxvcpus != 1)) { - virReportError(VIR_ERR_INTERNAL_ERROR, - _("Expecting domain XML entry 'vcpu' to be 1 or a " - "multiple of 2 but found %d"), - maxvcpus); + if (maxvcpus == 0) { + virReportError(VIR_ERR_INTERNAL_ERROR, "%s", + _("Expecting domain XML entry 'vcpu' to be greater " + "than 0")); goto cleanup; } -- 2.17.1

2 2

[libvirt] [PATCH] virDomainSnapshotDefParse: Prefer VIR_STEAL_PTR
by Michal Privoznik 22 Jun '18

22 Jun '18

Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- src/conf/snapshot_conf.c | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-) diff --git a/src/conf/snapshot_conf.c b/src/conf/snapshot_conf.c index 787c3d0feb..9c537ac7d1 100644 --- a/src/conf/snapshot_conf.c +++ b/src/conf/snapshot_conf.c @@ -331,8 +331,7 @@ virDomainSnapshotDefParse(xmlXPathContextPtr ctxt, "disk-only snapshot")); goto cleanup; } - def->file = memoryFile; - memoryFile = NULL; + VIR_STEAL_PTR(def->file, memoryFile); /* verify that memory path is absolute */ if (def->file && def->file[0] != '/') { @@ -372,7 +371,7 @@ virDomainSnapshotDefParse(xmlXPathContextPtr ctxt, if (!offline && virSaveCookieParse(ctxt, &def->cookie, saveCookie) < 0) goto cleanup; - ret = def; + VIR_STEAL_PTR(ret, def); cleanup: VIR_FREE(creation); @@ -380,8 +379,7 @@ virDomainSnapshotDefParse(xmlXPathContextPtr ctxt, VIR_FREE(nodes); VIR_FREE(memorySnapshot); VIR_FREE(memoryFile); - if (ret == NULL) - virDomainSnapshotDefFree(def); + virDomainSnapshotDefFree(def); return ret; } -- 2.16.4

2 1

[libvirt] [PATCH] qemu: Escape commas for qemuBuildDiskThrottling
by Anya Harter 22 Jun '18

22 Jun '18

Add comma escaping for disk->blkdeviotune.group_name. Signed-off-by: Anya Harter <aharter(a)redhat.com> --- src/qemu/qemu_command.c | 4 ++-- tests/qemuxml2argvdata/name-escape.args | 5 +++++ tests/qemuxml2argvdata/name-escape.xml | 13 +++++++++++++ tests/qemuxml2argvtest.c | 2 ++ 4 files changed, 22 insertions(+), 2 deletions(-) diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index 20c6ac2a04..e05b106a5e 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -1554,8 +1554,8 @@ qemuBuildDiskThrottling(virDomainDiskDefPtr disk, IOTUNE_ADD(size_iops_sec, "iops-size"); if (disk->blkdeviotune.group_name) { - virBufferEscapeString(buf, ",throttling.group=%s", - disk->blkdeviotune.group_name); + virBufferAddLit(buf, ",throttling.group="); + virQEMUBuildBufferEscapeComma(buf, disk->blkdeviotune.group_name); } IOTUNE_ADD(total_bytes_sec_max_length, "bps-total-max-length"); diff --git a/tests/qemuxml2argvdata/name-escape.args b/tests/qemuxml2argvdata/name-escape.args index 72ed2e8410..aef7c238ca 100644 --- a/tests/qemuxml2argvdata/name-escape.args +++ b/tests/qemuxml2argvdata/name-escape.args @@ -24,6 +24,11 @@ bar=2/monitor.sock,server,nowait \ -boot c \ -device usb-ccid,id=ccid0,bus=usb.0,port=1 \ -usb \ +-drive file=/dev/HostVG/QEMUGuest1,format=qcow2,if=none,id=drive-ide0-0-0,\ +cache=none,throttling.bps-total=5000,throttling.iops-total=6000,\ +throttling.bps-total-max=10000,throttling.iops-total-max=11000,\ +throttling.group=libvirt_iotune_group1,,foo \ +-device ide-drive,bus=ide.0,unit=0,drive=drive-ide0-0-0,id=ide0-0-0 \ -device ccid-card-emulated,backend=certificates,cert1=cert1,,foo,cert2=cert2,\ cert3=cert3,db=/etc/pki/nssdb,,foo,id=smartcard0,bus=ccid0.0 \ -chardev tty,id=charserial0,path=/dev/ttyS2,,foo \ diff --git a/tests/qemuxml2argvdata/name-escape.xml b/tests/qemuxml2argvdata/name-escape.xml index 0580de1813..70a1ce09d3 100644 --- a/tests/qemuxml2argvdata/name-escape.xml +++ b/tests/qemuxml2argvdata/name-escape.xml @@ -14,6 +14,19 @@ <on_crash>destroy</on_crash> <devices> <emulator>/usr/bin/qemu-system-i686</emulator> + <disk type='block' device='disk'> + <driver name='qemu' type='qcow2' cache='none'/> + <source dev='/dev/HostVG/QEMUGuest1'/> + <target dev='hda' bus='ide'/> + <iotune> + <total_bytes_sec>5000</total_bytes_sec> + <total_iops_sec>6000</total_iops_sec> + <total_bytes_sec_max>10000</total_bytes_sec_max> + <total_iops_sec_max>11000</total_iops_sec_max> + <group_name>libvirt_iotune_group1,foo</group_name> + </iotune> + <address type='drive' controller='0' bus='0' target='0' unit='0'/> + </disk> <graphics type='vnc'> <listen type='socket'/> </graphics> diff --git a/tests/qemuxml2argvtest.c b/tests/qemuxml2argvtest.c index a9a493e308..582a9de7bb 100644 --- a/tests/qemuxml2argvtest.c +++ b/tests/qemuxml2argvtest.c @@ -2758,6 +2758,8 @@ mymain(void) DO_TEST("name-escape", QEMU_CAPS_NAME_DEBUG_THREADS, QEMU_CAPS_OBJECT_SECRET, + QEMU_CAPS_DRIVE_IOTUNE_MAX, + QEMU_CAPS_DRIVE_IOTUNE_GROUP, QEMU_CAPS_VNC, QEMU_CAPS_NAME_GUEST, QEMU_CAPS_DEVICE_CIRRUS_VGA, -- 2.17.1

2 1

[libvirt] [PATCH v3 0/4] cmdDomblkinfo: introduce --all to show all block devices info
by Chen Hanxiao 22 Jun '18

22 Jun '18

This series introduce --all to cmdDomblkinfo to show all block devices info in one cmd. Likes a combination of domblklist and domblkinfo. v3: check error code on network disk v2: add support --human for --all v1.1 fix a self test Chen Hanxiao (4): cmdDomblkinfo: introduce helper cmdDomblkinfoPrint cmdDomblkinfo: add --all to show all block devices info cmdDomblkinfoPrint: support printing "-" for invalid virDomainBlockInfo news: add cmdDomblkinfo --all option docs/news.xml | 10 +++ tools/virsh-domain-monitor.c | 160 ++++++++++++++++++++++++++++++----- tools/virsh.pod | 5 +- 3 files changed, 155 insertions(+), 20 deletions(-) -- 2.17.1

2 7

[libvirt] ANNOUNCE: libvirt-dbus 1.1.0 released
by Pavel Hrdina 21 Jun '18

21 Jun '18

I'm happy to announce the release of libvirt-dbus 1.1.0. libvirt-dbus wraps libvirt API to provide high-level object-oriented API better suited for dbus-based applications. You can download it here: https://libvirt.org/sources/dbus/libvirt-dbus-1.1.0.tar.gz * New features - Support for all relevant nwfilter APIs up to libvirt 3.0.0 - Support for all relevant storage volume APIs up to libvirt 3.0.0 - Support for all relevant node device APIs up to libvirt 3.0.0 * Bug fixes - Don't report error for GetAll on properties if some property is not accessible Thanks everybody who contributed! Pavel

1 0

[libvirt] [jenkins-ci PATCH 0/2] git changes to reflect new CI configuration
by Pavel Hrdina 21 Jun '18

21 Jun '18

Pavel Hrdina (2): jobs: Change git URL to github jobs: Switch to github webhook instead of polling jobs/autotools.yaml | 5 ++--- jobs/generic.yaml | 5 ++--- jobs/go.yaml | 5 ++--- jobs/perl-modulebuild.yaml | 5 ++--- jobs/python-distutils.yaml | 5 ++--- 5 files changed, 10 insertions(+), 15 deletions(-) -- 2.17.1

3 6