May 2013 - Devel - libvirt List Archives

[libvirt] two hostdev devices problem
by Dominik Mostowiec 27 May '13

27 May '13

hi, I try to add 2 VF by "hostdev". Networks (vnet0, vnet1) with: <forward mode='hostdev' managed='yes'> <pf dev='eth1'/> ..... Domain: <interface type="network"> <source network="vnet0"/> .... <interface type="network"> <source network="vnet1"/> .... virsh create error: "error: internal error process exited while connecting to monitor: kvm: -device pci-assign,configfd=25,host=01:10.1,id=hostdev0,bus=pci.0,addr=0x4: Duplicate ID 'hostdev0' for device" I am doing somthing wrong? My configs: domain: <domain xmlns:qemu="http://libvirt.org/schemas/domain/qemu/1.0" type="kvm"> <name>one-466</name> <vcpu>1</vcpu> <memory>1781760</memory> <os> <type arch="x86_64">hvm</type> <boot dev="hd"/> </os> <devices> <emulator>/usr/bin/kvm</emulator> <disk type="file" device="disk"> <source file="/nebula/nebula/var//datastores/0/466/disk.0"/> <target dev="hda" bus="virtio"/> <driver name="qemu" type="raw" cache="default"/> </disk> <disk type="file" device="disk"> <source file="/nebula/nebula/var//datastores/0/466/disk.1"/> <target dev="vdb" bus="virtio"/> <driver name="qemu" type="raw" cache="default"/> </disk> <graphics type="vnc" listen="0.0.0.0" port="-1"/> <interface type="network"> <source network="vnet0"/> <mac address="52:54:0a:b1:48:91"/> </interface> <interface type="network"> <source network="vnet1"/> <mac address="52:54:0a:b1:48:92"/> </interface> </devices> <features> <acpi/> </features> <devices><serial type="pty"><source path="/dev/pts/477"/><target port="0"/></serial><console type="pty" tty="/dev/pts/477"><source path="/dev/pts/477"/><target port="0"/></console></devices> </domain> Networks: cat /etc/libvirt/qemu/networks/vnet0.xml  <network> <name>vnet0</name> <uuid>1b6a3506-d9e0-4e1e-bbff-6f700b8dc80b</uuid> <forward mode='hostdev' managed='yes'> <pf dev='eth0'/> </forward> </network> cat /etc/libvirt/qemu/networks/vnet1.xml  <network> <name>vnet1</name> <uuid>11feb088-d609-4177-86d7-ac3b9467a3e8</uuid> <forward mode='hostdev' managed='yes'> <pf dev='eth1'/> </forward> </network> I wrote libvirt-users mailing list too. -- Pozdrawiam Dominik

4 15

[libvirt] [test-API][PATCH] Add 2 emulatorpin cases cover config and live flags
by Wayne Sun 27 May '13

27 May '13

- use pinEmulator to pin domain emulator to host cpu - 2 cases cover config and live flags - cpulist with '^', '-' and ',' is supported to give multiple host cpus Related bug 916493: pinEmulator and emulatorPinInfo should be simple with required params https://bugzilla.redhat.com/show_bug.cgi?id=916493 is fixed, so the test can run successfully now. Signed-off-by: Wayne Sun <gsun(a)redhat.com> --- repos/setVcpus/emulatorpin_config.py | 97 +++++++++++++++++++++++++++++++++++ repos/setVcpus/emulatorpin_live.py | 98 ++++++++++++++++++++++++++++++++++++ 2 files changed, 195 insertions(+) create mode 100644 repos/setVcpus/emulatorpin_config.py create mode 100644 repos/setVcpus/emulatorpin_live.py diff --git a/repos/setVcpus/emulatorpin_config.py b/repos/setVcpus/emulatorpin_config.py new file mode 100644 index 0000000..9b94f98 --- /dev/null +++ b/repos/setVcpus/emulatorpin_config.py @@ -0,0 +1,97 @@ +#!/usr/bin/env python +# Test domain emulator pin with flag VIR_DOMAIN_AFFECT_CONFIG, check +# domain config xml with emulatorpin configuration. + +import re +from xml.dom import minidom + +import libvirt +from libvirt import libvirtError + +from src import sharedmod +from utils import utils + +required_params = ('guestname', 'cpulist',) +optional_params = {} + +def emulatorpin_check(domobj, cpumap): + """check domain config xml with emulatorpin element + """ + guestxml = domobj.XMLDesc(2) + logger.debug("domain %s xml :\n%s" %(domobj.name(), guestxml)) + + doc = minidom.parseString(guestxml) + emulatorpin = doc.getElementsByTagName('emulatorpin') + if not emulatorpin: + logger.error("no emulatorpin element in domain xml") + return 1 + + if not emulatorpin[0].hasAttribute('cpuset'): + logger.error("no cpuset attribute with emulatorpin in domain xml") + return 1 + else: + emulator_attr = emulatorpin[0].getAttributeNode('cpuset') + cpulist = emulator_attr.nodeValue + cpumap_tmp = utils.param_to_tuple(cpulist, maxcpu) + + if cpumap_tmp == cpumap: + logger.info("cpuset is as expected in domain xml") + return 0 + else: + logger.error("cpuset is not as expected in domain xml") + return 1 + +def emulatorpin_config(params): + """pin domain emulator to host cpu with config flag + """ + global logger + logger = params['logger'] + params.pop('logger') + guestname = params['guestname'] + cpulist = params['cpulist'] + + logger.info("the name of virtual machine is %s" % guestname) + logger.info("the given cpulist is %s" % cpulist) + + global maxcpu + maxcpu = utils.get_host_cpus() + logger.info("%s physical cpu on host" % maxcpu) + + conn = sharedmod.libvirtobj['conn'] + + try: + domobj = conn.lookupByName(guestname) + cpumap = utils.param_to_tuple(cpulist, maxcpu) + + if not cpumap: + logger.error("cpulist: Invalid format") + return 1 + + logger.debug("cpumap for emulator pin is:") + logger.debug(cpumap) + + logger.info("pin domain emulator to host cpulist %s with flag: %s" % + (cpulist, libvirt.VIR_DOMAIN_AFFECT_CONFIG)) + domobj.pinEmulator(cpumap, libvirt.VIR_DOMAIN_AFFECT_CONFIG) + + logger.info("check emulator pin info") + ret = domobj.emulatorPinInfo(libvirt.VIR_DOMAIN_AFFECT_CONFIG) + logger.debug("emulator pin info is:") + logger.debug(ret) + if ret == cpumap: + logger.info("emulator pin info is expected") + else: + logger.error("emulator pin info is not expected") + return 1 + except libvirtError, e: + logger.error("libvirt call failed: " + str(e)) + return 1 + + logger.info("check domain emulatorpin configuration in xml") + ret = emulatorpin_check(domobj, cpumap) + if ret: + logger.error("domain emulator pin check failed") + return 1 + else: + logger.info("domain emulator pin check succeed") + return 0 diff --git a/repos/setVcpus/emulatorpin_live.py b/repos/setVcpus/emulatorpin_live.py new file mode 100644 index 0000000..08b7073 --- /dev/null +++ b/repos/setVcpus/emulatorpin_live.py @@ -0,0 +1,98 @@ +#!/usr/bin/env python +# Test domain emulator pin with flag VIR_DOMAIN_AFFECT_LIVE, check +# emulator process status under domain task list on host. + +import re + +import libvirt +from libvirt import libvirtError + +from src import sharedmod +from utils import utils + +required_params = ('guestname', 'cpulist',) +optional_params = {} + +def emulatorpin_check(guestname, cpumap): + """check emulator process status of the running virtual machine + grep Cpus_allowed_list /proc/PID/status + """ + tmp_str = '' + cmd = "cat /var/run/libvirt/qemu/%s.pid" % guestname + status, pid = utils.exec_cmd(cmd, shell=True) + if status: + logger.error("failed to get the pid of domain %s" % guestname) + return 1 + + cmd = "grep Cpus_allowed_list /proc/%s/status" % pid[0] + status, output = utils.exec_cmd(cmd, shell=True) + if status: + logger.error("failed to run command: %s" % cmd) + return 1 + + logger.debug("command '%s' output is:\n%s" % (cmd, output[0])) + + cpulist = output[0].split('\t')[1] + ret = utils.param_to_tuple(cpulist, maxcpu) + + if ret == cpumap: + logger.info("emulator process cpus allowed list is expected") + return 0 + else: + logger.error("emulator process cpus allowed list is not expected") + return 1 + +def emulatorpin_live(params): + """pin domain emulator to host cpu with live flag + """ + global logger + logger = params['logger'] + params.pop('logger') + guestname = params['guestname'] + cpulist = params['cpulist'] + + logger.info("the name of virtual machine is %s" % guestname) + logger.info("the given cpulist is %s" % cpulist) + + global maxcpu + maxcpu = utils.get_host_cpus() + logger.info("%s physical cpu on host" % maxcpu) + + conn = sharedmod.libvirtobj['conn'] + + try: + domobj = conn.lookupByName(guestname) + cpumap = utils.param_to_tuple(cpulist, maxcpu) + if not cpumap: + logger.error("cpulist: Invalid format") + return 1 + + logger.debug("cpumap for emulator pin is:") + logger.debug(cpumap) + + logger.info("pin domain emulator to host cpu %s with flag: %s" % + (cpulist, libvirt.VIR_DOMAIN_AFFECT_LIVE)) + domobj.pinEmulator(cpumap, libvirt.VIR_DOMAIN_AFFECT_LIVE) + + logger.info("check emulator info") + ret = domobj.emulatorPinInfo() + logger.debug("emulator info is:") + logger.debug(ret) + if ret == cpumap: + logger.info("emulator info is expected") + else: + logger.error("emulator info is not expected") + return 1 + + except libvirtError, e: + logger.error("libvirt call failed: " + str(e)) + return 1 + + logger.info("check emulator pin status on host") + ret = emulatorpin_check(guestname, cpumap) + if ret: + logger.error("domain emulator pin failed") + return 1 + else: + logger.info("domain emulator pin succeed") + return 0 -- 1.8.2.1

1 0

[libvirt] [PATCH] qemuOpenVhostNet: Decrease vhostfdSize on open failure
by Michal Privoznik 27 May '13

27 May '13

Currently, if there's an error opening /dev/vhost-net (e.g. because it doesn't exist) but it's not required we proceed with vhostfd array filled with -1 and vhostfdSize unchanged. Later, when constructing the qemu command line only non-negative items within vhostfd array are taken into account. This means, vhostfdSize may be greater than the actual count of non-negative items in vhostfd array. This results in improper command line arguments being generated, e.g.: -netdev tap,fd=21,id=hostnet0,vhost=on,vhostfd=(null) --- src/qemu/qemu_command.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index 434f5a7..d969f88 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -486,6 +486,8 @@ qemuOpenVhostNet(virDomainDefPtr def, "but is unavailable")); goto error; } + i--; + (*vhostfdSize)--; } } virDomainAuditNetDevice(def, net, "/dev/vhost-net", *vhostfdSize); @@ -6560,12 +6562,10 @@ qemuBuildInterfaceCommandLine(virCommandPtr cmd, } for (i = 0; i < vhostfdSize; i++) { - if (vhostfd[i] >= 0) { - virCommandTransferFD(cmd, vhostfd[i]); - if (virAsprintf(&vhostfdName[i], "%d", vhostfd[i]) < 0) { - virReportOOMError(); - goto cleanup; - } + virCommandTransferFD(cmd, vhostfd[i]); + if (virAsprintf(&vhostfdName[i], "%d", vhostfd[i]) < 0) { + virReportOOMError(); + goto cleanup; } } -- 1.8.2.1

2 2

[libvirt] [PATCH 0/2] Fix the display of *virsh vcpuinfo* command.
by yangdongsheng 26 May '13

26 May '13

Hi, I found some confusion in display of commad *virsh vcpuinfo domname* . There is no diffrence between a cpu offline and a cpu online but not be set for the affinity of vcpu, both are '-'. But users will find the difference of them in using them for a same action (such as vcpupin) and be confused. Example: [root@yds-pc ~]# echo "0" >/sys/devices/system/cpu/cpu2/online [root@yds-pc ~]# cat /proc/cpuinfo |grep processor processor : 0 processor : 1 processor : 3 [root@yds-pc ~]# virsh vcpupin virt-tests-vm1 0 0,3 [root@yds-pc ~]# virsh vcpuinfo virt-tests-vm1 VCPU: 0 CPU: 0 State: running CPU time: 14.8s CPU Affinity: y--y *There is no diff between cpu1 and cpu2 (both '-')* BUT: [root@yds-pc ~]# virsh vcpupin virt-tests-vm1 0 1 [root@yds-pc ~]# echo $? 0 [root@yds-pc ~]# virsh vcpupin virt-tests-vm1 0 2 error: Requested operation is not valid: failed to set cpuset.cpus in cgroup for vcpu 0 [root@yds-pc ~]# echo $? 1 So, I think we should show the cpu in different state with different symbol. In this patchset, I introduce a symbol 'x' (maybe there is a better one :)) to display the cpu is offline on host. Just like: [root@yds-pc ~]# virsh vcpuinfo virt-tests-vm1 VCPU: 0 CPU: 0 State: running CPU time: 14.8s CPU Affinity: y-xy yangdongsheng (2): tool/virsh-domain.c: Add a vshNodeGetCPUMap function to get cpumap of host. tool/virsh-domain.c: Fix the display of Affinity in function cmdVcpuinfo. tools/virsh-domain.c | 78 ++++++++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 75 insertions(+), 3 deletions(-) -- 1.7.10.1

4 6

[libvirt] virCommandWait internal error Child process unexpected exit status 1
by jj 25 May '13

25 May '13

I am use libvirt master branch to test LXC , the lastest commit of the master branch is 03d813bbcd7b4a18360105500672b84d985dd889, by Marek Marczykowski <marmarek(a)invisiblethingslab.com> on Thu May 23 02:01:30 2013 +0200, root@wheezy:/tmp# cat test.xml <domain type='lxc'> <name>vmtest</name> <memory>500000</memory> <os> <type>exe</type> <cmdline>console=ttyS0</cmdline> <init>/sbin/init</init> </os> <vcpu>1</vcpu> <clock offset='utc'/> <on_poweroff>destroy</on_poweroff> <on_reboot>restart</on_reboot> <on_crash>destroy</on_crash> <devices> <filesystem type="mount"> <source dir="/tmp/x"/> <target dir="/"/> </filesystem> <emulator>/usr/libexec/libvirt_lxc</emulator> <interface type='network'> <source network='default'/> </interface> <console type='pty' /> </devices> </domain> root@wheezy:/tmp# virsh -c lxc:/// define test.xml Domain vmtest defined from test.xml root@wheezy:/tmp# virsh -c lxc:/// start vmtest error: Failed to start domain vmtest error: internal error Child process (PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin LIBVIRT_DEBUG=3 LIBVIRT_LOG_OUTPUTS=3:stderr /usr/libexec/libvirt_lxc --name vmtest --console 19 --security=none --handshake 22 --background --veth veth7) unexpected exit status 1 below's a libvirtd log 2013-05-25 14:46:58.727+0000: 7865: error : virCommandWait:2354 : internal error Child process (PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin LIBVIRT_DEBUG=3 LIBVIRT_LOG_OUTPUTS=3:stderr /usr/libexec/libvirt_lxc --name vmtest --console 19 --security=none --handshake 22 --background --veth veth7) unexpected exit status 1

1 0

Re: [libvirt] [Qemu-devel] [PATCH] qapi-schema.json: Reformat TargetType enum to one-per-line
by Anthony Liguori 25 May '13

25 May '13

Peter Maydell <peter.maydell(a)linaro.org> writes: > On 22 May 2013 14:15, Anthony Liguori <aliguori(a)us.ibm.com> wrote: >> Paolo Bonzini <pbonzini(a)redhat.com> writes: >>> You >>> don't need to know what targets were supported in the version that you >>> compiled from. Only one target is supported in this executable >>> anyway. >> >> It seems useful to me. One day we may support multiple targets per >> executable. > > Why would you care about which architectures the executable supports? > What you actually want to know is which machine models are supported; > whether board foo happens to be ARM or PPC isn't really very interesting > IMHO. That's a very good point. It was the libvirt folks that requested this. Perhaps they can shed some light on the logic? Regards, Anthony Liguori > > -- PMM

5 4

[libvirt] [PATCH] qemu: save domain state to XML after reboot
by root 24 May '13

24 May '13

Currently qemuDomainReboot() does reboot in two phases: qemuMonitorSystemPowerdown() and qemuProcessFakeReboot(). qemuMonitorSystemPowerdown() shutdowns the domain and saves domain state/reason as VIR_DOMAIN_SHUTDOWN_UNKNOWN. qemuProcessFakeReboot() sets domain state/reason to VIR_DOMAIN_RESUMED_UNPAUSED but does not save domain state changes. Subsequent restart of libvirtd leads to restoring domain state/reason to saved that is VIR_DOMAIN_SHUTDOWN_UNKNOWN and to automatic shutdown of the domain. This commit adds virDomainSaveStatus() into qemuProcessFakeReboot() to avoid unexpected shutdowns. --- src/qemu/qemu_process.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index e5b4679..cbe35ac 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -547,6 +547,7 @@ qemuProcessFakeReboot(void *opaque) virDomainObjPtr vm = opaque; qemuDomainObjPrivatePtr priv = vm->privateData; virDomainEventPtr event = NULL; + virQEMUDriverConfigPtr cfg = virQEMUDriverGetConfig(driver); int ret = -1; VIR_DEBUG("vm=%p", vm); virObjectLock(vm); @@ -585,6 +586,11 @@ qemuProcessFakeReboot(void *opaque) VIR_DOMAIN_EVENT_RESUMED, VIR_DOMAIN_EVENT_RESUMED_UNPAUSED); + if (virDomainSaveStatus(driver->xmlopt, cfg->stateDir, vm) < 0) { + VIR_WARN("Unable to save status on vm %s after state change", + vm->def->name); + } + ret = 0; endjob: @@ -601,6 +607,7 @@ cleanup: } if (event) qemuDomainEventQueue(driver, event); + virObjectUnref(cfg); } -- 1.8.2.1

3 2

[libvirt] virStream double unref in virChrdevOpen()
by Sergey Fionov 24 May '13

24 May '13

Hello, There is double unref virChrdevOpen() (src/conf/virchrdev.c) when error occured. if (virStreamRef(st) < 0) { virMutexUnlock(&devs->lock); return -1; } ... if (virHashAddEntry(devs->hash, path, st) < 0) goto error; ... if (virFDStreamOpenFile(st, path, 0, 0, O_RDWR) < 0) /* error occured here */ goto error; ... error: virStreamFree(st); virHashRemoveEntry(devs->hash, path); stream is virStreamRef'ed 1 time but if it is successfully placed into hash then it will be unreferenced 2 times - in virStreamFree() and virHashRemoveEntry()'s dataFree callback. That leads to dispose stream object and segmentation fault due to use after free. Steps to reproduce: # hide the /dev/pts to throw an error in virFDStreamOpenFile() $ mount -t tmpfs empty-devpts /dev/pts $ virsh console a111 Connected to domain a111 Escape character is ^] error: End of file while reading data: Input/output error error: One or more references were leaked after disconnect from the hypervisor error: Failed to reconnect to the hypervisor (libvirtd segfaults) $ umount empty-devpts wbr, Sergey.

2 1

[libvirt] [PATCH] esx: Fix dynamic VI object type detection
by Matthias Bolte 24 May '13

24 May '13

VI objects support inheritance with subtype polymorphism. For example the FileInfo object type is extended by FloppyImageFileInfo, FolderFileInfo etc. Then SearchDatastore_Task returns an array of FileInfo objects and depending on the represented file the FileInfo is actually a FolderFileInfo or FloppyImageFileInfo etc. The actual type information is stored as XML attribute that allows clients such as libvirt to distinguish between the actual types. esxVI_GetActualObjectType is used to extract the actual type. I assumed that this mechanism would be used for all VI object types that have subtypes. But this is not the case. It seems only to be used for types that are actually used as generic base type such as FileInfo. But it is not used for types that got extended later such as ElementDescription that was extended by ExtendedElementDescription (added in vSphere API 4.0) or that are not meant to be used with subtype polymorphism. This breaks the deserialization of types that contain ElementDescription properties such as PerfCounterInfo or ChoiceOption, because the code expects an ElementDescription object to have an XML attribute named type that is not present, since ExtendedElementDescription was added to the esx_vi_generator.input in commit 60f0f55ee4686fecbffc5fb32f90863427d02a14. This in turn break virtual machine question handling and auto answering. Fix this by using the base type if no XML type attribute is present. --- src/esx/esx_vi_types.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/src/esx/esx_vi_types.c b/src/esx/esx_vi_types.c index 1a26556..c55396c 100644 --- a/src/esx/esx_vi_types.c +++ b/src/esx/esx_vi_types.c @@ -720,10 +720,9 @@ esxVI_GetActualObjectType(xmlNodePtr node, esxVI_Type baseType, BAD_CAST "http://www.w3.org/2001/XMLSchema-instance"); if (type == NULL) { - virReportError(VIR_ERR_INTERNAL_ERROR, - _("%s is missing 'type' property"), - esxVI_Type_ToString(baseType)); - return -1; + // no actual type specified, use base type instead + *actualType = baseType; + return 0; } *actualType = esxVI_Type_FromString(type); -- 1.7.9.5

2 2

[libvirt] [PATCH] spec: Build vbox packages only for x86 architectures
by Viktor Mihajlovski 24 May '13

24 May '13

Commit 6ab6bc19f03513fd87d29ecfd405bb7f4a7de114 has introduced separate daemon/driver packages for vbox. These should only be built for x86 architectures which is done hereby. Signed-off-by: Viktor Mihajlovski <mihajlov(a)linux.vnet.ibm.com> --- libvirt.spec.in | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/libvirt.spec.in b/libvirt.spec.in index e1c5f49..7abe9fe 100644 --- a/libvirt.spec.in +++ b/libvirt.spec.in @@ -130,6 +130,11 @@ %define with_libxl 0 %endif +# vbox is available only on i386 x86_64 +%ifnarch %{ix86} x86_64 + %define with_vbox 0 +%endif + # Numactl is not available on s390[x] and ARM %ifarch s390 s390x %{arm} %define with_numactl 0 -- 1.7.9.5

2 1