August 2016 - Devel - libvirt List Archives

[libvirt] [PATCH 0/4] Fix host-model CPUs on hosts with CMT
by Jiri Denemark 10 Aug '16

10 Aug '16

Since the introduction of CMT features (commit v1.3.5-461-gf294b83) starting a domain with host-model CPU on a host which supports CMT fails because QEMU complains about unknown 'cmt' feature: qemu-system-x86_64: CPU feature cmt not found https://bugzilla.redhat.com/show_bug.cgi?id=1355857 Jiri Denemark (4): cpu_x86: Introduce x86FeatureIsMigratable cpu_x86: Properly drop non-migratable features tests: Add a test for host-model CPU with CMT feature cpu_x86: Fix host-model CPUs on hosts with CMT src/cpu/cpu_x86.c | 53 +++++++++++++++------- .../qemuxml2argv-cpu-host-model-cmt.args | 22 +++++++++ .../qemuxml2argv-cpu-host-model-cmt.xml | 19 ++++++++ tests/qemuxml2argvtest.c | 1 + tests/testutilsqemu.c | 1 + 5 files changed, 79 insertions(+), 17 deletions(-) create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-cpu-host-model-cmt.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-cpu-host-model-cmt.xml -- 2.9.2

2 6

[libvirt] [PATCH] virsh: clarify snapshot --live
by Chen Hanxiao 10 Aug '16

10 Aug '16

From: Chen Hanxiao <chenhanxiao(a)gmail.com> In libvirt, snapshot means disk snapshot. snapshot --live is more like VM checkpoint. Make it clear in virsh.pod. Signed-off-by: Chen Hanxiao <chenhanxiao(a)gmail.com> --- tools/virsh.pod | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/tools/virsh.pod b/tools/virsh.pod index fc6a680..30e6bf3 100644 --- a/tools/virsh.pod +++ b/tools/virsh.pod @@ -3896,8 +3896,9 @@ this. If this flag is not specified, then some hypervisors may fail after partially performing the action, and B<dumpxml> must be used to see whether any partial changes occurred. -If I<--live> is specified, libvirt takes the snapshot while the guest is -running. This increases the size of the memory image of the external +If I<--live> is specified, libvirt takes the snapshot(checkpoint) while +the guest is running. Both disk snapshot and VM memory snapshot will be took. +This increases the size of the memory image of the external checkpoint. This is currently supported only for external checkpoints. Existence of snapshot metadata will prevent attempts to B<undefine> -- 1.8.3.1

2 2

[libvirt] [RFC] Design executing commands from within domains
by Michal Privoznik 10 Aug '16

10 Aug '16

Dear list, while wiring qemu-ga into libvirt I've noticed that it has ability to spawn commands inside guest. I haven't paid much attention to it then as implementing libvirt <-> qemu-ga communication was more important. But lately couple of requests on the list showed up where ability to spawn various commands inside guests would be much appreciated (e.g. when fetching some stats that HV can't know or has no support for yet - free/df/..). When it comes to spawning commands we distinguish two modes: sync and async. I think we should stick to this on the public API level too. Yet better, we can have async APIs and the sync API would just be a wrapper around async then. One picture instead of thousand words: /* obtain conn and dom somehow */ const char argv[] = {"argv1", "argv2", "argv3", ...} size_t argc = ARRAY_CARDINALITY(argv); unsigned int flags = 0; virDomainCommandPtr cmd = virDomainCommandNew("/path/to/binary", argv, argc, flags); /* Should we care about env too? */ virStreamPtr st = virSreamNew(conn, VIR_STREAM_NONBLOCK); virDomainCommandRun(dom, cmd, st, VIR_COMMAND_STRING_IO); virStreamEventAddCallback(st, VIR_STREAM_EVENT_READABLE, domainCommandCallback, ...); /* ... */ virDomainCommandJoin(dom, cmd); virStreamEventRemoveCallback(con->st) virDomainCommandFree(cmd); While qemu-ga can handle input for a binary its executing, this must be put right into JSON when constructing the qemu-ga command. It's not common that by that time users know the input they want to enter. So I have to figure out something. Opinions? Michal

4 9

[libvirt] [PATCH 0/3] Adjust device compat. rules for dmi-to-pci-bridge and pcie-bus-expander
by Laine Stump 10 Aug '16

10 Aug '16

I got a few things wrong in the original implementation. Details in the individual patches. Laine Stump (3): conf: improve error log when PCI devices don't match requested controller conf: don't allow connecting upstream-port directly to pce-expander-bus conf: restrict where dmi-to-pci-bridge can be connected docs/formatdomain.html.in | 24 ++++--- src/conf/domain_addr.c | 79 ++++++++++++---------- src/conf/domain_addr.h | 4 +- .../qemuxml2argv-q35-dmi-bad-address1.xml | 39 +++++++++++ .../qemuxml2argv-q35-dmi-bad-address2.xml | 39 +++++++++++ tests/qemuxml2argvtest.c | 16 +++++ 6 files changed, 155 insertions(+), 46 deletions(-) create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-q35-dmi-bad-address1.xml create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-q35-dmi-bad-address2.xml -- 2.7.4

2 12

[libvirt] [PATCH v3] virsh: Introduce nodedev-event command
by Jovanka Gulicoska 10 Aug '16

10 Aug '16

Add nodedev-event support for node device lifecycle events --- tools/virsh-nodedev.c | 187 ++++++++++++++++++++++++++++++++++++++++++++++++++ tools/virsh.pod | 18 +++++ 2 files changed, 205 insertions(+) diff --git a/tools/virsh-nodedev.c b/tools/virsh-nodedev.c index a715b61..d3d3fc1 100644 --- a/tools/virsh-nodedev.c +++ b/tools/virsh-nodedev.c @@ -31,6 +31,7 @@ #include "viralloc.h" #include "virfile.h" #include "virstring.h" +#include "virtime.h" #include "conf/node_device_conf.h" /* @@ -739,6 +740,186 @@ cmdNodeDeviceReset(vshControl *ctl, const vshCmd *cmd) return ret; } +/* + * "nodedev-event" command + */ +VIR_ENUM_DECL(virshNodeDeviceEvent) +VIR_ENUM_IMPL(virshNodeDeviceEvent, + VIR_NODE_DEVICE_EVENT_LAST, + N_("Created"), + N_("Deleted")) + +static const char * +virshNodeDeviceEventToString(int event) +{ + const char *str = virshNodeDeviceEventTypeToString(event); + return str ? _(str) : _("unknown"); +} + +struct virshNodeDeviceEventData { + vshControl *ctl; + bool loop; + bool timestamp; + int count; +}; +typedef struct virshNodeDeviceEventData virshNodeDeviceEventData; + +VIR_ENUM_DECL(virshNodeDeviceEventId) +VIR_ENUM_IMPL(virshNodeDeviceEventId, + VIR_NODE_DEVICE_EVENT_ID_LAST, + "lifecycle") + +static void +vshEventLifecyclePrint(virConnectPtr conn ATTRIBUTE_UNUSED, + virNodeDevicePtr dev, + int event, + int detail ATTRIBUTE_UNUSED, + void *opaque) +{ + virshNodeDeviceEventData *data = opaque; + + if (!data->loop && data->count) + return; + + if (data->timestamp) { + char timestamp[VIR_TIME_STRING_BUFLEN]; + + if (virTimeStringNowRaw(timestamp) < 0) + timestamp[0] = '\0'; + + vshPrint(data->ctl, _("%s: event 'lifecycle' for node device %s: %s\n"), + timestamp, + virNodeDeviceGetName(dev), virshNodeDeviceEventToString(event)); + } else { + vshPrint(data->ctl, _("event 'lifecycle' for node device %s: %s\n"), + virNodeDeviceGetName(dev), virshNodeDeviceEventToString(event)); + } + + data->count++; + if (!data->loop) + vshEventDone(data->ctl); +} + +static const vshCmdInfo info_node_device_event[] = { + {.name = "help", + .data = N_("Node Device Events") + }, + {.name = "desc", + .data = N_("List event types, or wait for node device events to occur") + }, + {.name = NULL} +}; + +static const vshCmdOptDef opts_node_device_event[] = { + {.name = "device", + .type = VSH_OT_STRING, + .help = N_("filter by node device name") + }, + {.name = "event", + .type = VSH_OT_STRING, + .help = N_("which event type to wait for") + }, + {.name = "loop", + .type = VSH_OT_BOOL, + .help = N_("loop until timeout or interrupt, rather than one-shot") + }, + {.name = "timeout", + .type = VSH_OT_INT, + .help = N_("timeout seconds") + }, + {.name = "list", + .type = VSH_OT_BOOL, + .help = N_("list valid event types") + }, + {.name = "timestamp", + .type = VSH_OT_BOOL, + .help = N_("show timestamp for each printed event") + }, + {.name = NULL} +}; + +static bool +cmdNodeDeviceEvent(vshControl *ctl, const vshCmd *cmd) +{ + virNodeDevicePtr dev = NULL; + bool ret = false; + int eventId = -1; + int timeout = 0; + virshNodeDeviceEventData data; + const char *eventName = NULL; + const char *device_value = NULL; + int event; + virshControlPtr priv = ctl->privData; + + if (vshCommandOptBool(cmd, "list")) { + size_t i; + + for (i = 0; i < VIR_NODE_DEVICE_EVENT_ID_LAST; i++) + vshPrint(ctl, "%s\n", virshNodeDeviceEventIdTypeToString(i)); + return true; + } + + if (vshCommandOptStringReq(ctl, cmd, "event", &eventName) < 0) + return false; + if (!eventName) { + vshError(ctl, "%s", _("either --list or event type is required")); + return false; + } + if ((event = virshNodeDeviceEventIdTypeFromString(eventName)) < 0) { + vshError(ctl, _("unknown event type %s"), eventName); + return false; + } + + data.ctl = ctl; + data.loop = vshCommandOptBool(cmd, "loop"); + data.timestamp = vshCommandOptBool(cmd, "timestamp"); + data.count = 0; + if (vshCommandOptTimeoutToMs(ctl, cmd, &timeout) < 0) + return false; + if (vshCommandOptStringReq(ctl, cmd, "device", &device_value) < 0) + return false; + + if (device_value) { + if (!(dev = virNodeDeviceLookupByName(priv->conn, device_value))) { + vshError(ctl, "%s '%s'", + _("Could not find matching device"), device_value); + goto cleanup; + } + } + if (vshEventStart(ctl, timeout) < 0) + goto cleanup; + + if ((eventId = virConnectNodeDeviceEventRegisterAny(priv->conn, dev, event, + VIR_NODE_DEVICE_EVENT_CALLBACK(vshEventLifecyclePrint), + &data, NULL)) < 0) + goto cleanup; + switch (vshEventWait(ctl)) { + case VSH_EVENT_INTERRUPT: + vshPrint(ctl, "%s", _("event loop interrupted\n")); + break; + case VSH_EVENT_TIMEOUT: + vshPrint(ctl, "%s", _("event loop timed out\n")); + break; + case VSH_EVENT_DONE: + break; + default: + goto cleanup; + } + vshPrint(ctl, _("events received: %d\n"), data.count); + if (data.count) + ret = true; + + cleanup: + vshEventCleanup(ctl); + if (eventId >= 0 && + virConnectNodeDeviceEventDeregisterAny(priv->conn, eventId) < 0) + ret = false; + if (dev) + virNodeDeviceFree(dev); + return ret; +} + + const vshCmdDef nodedevCmds[] = { {.name = "nodedev-create", .handler = cmdNodeDeviceCreate, @@ -788,5 +969,11 @@ const vshCmdDef nodedevCmds[] = { .info = info_node_device_reset, .flags = 0 }, + {.name = "nodedev-event", + .handler = cmdNodeDeviceEvent, + .opts = opts_node_device_event, + .info = info_node_device_event, + .flags = 0 + }, {.name = NULL} }; diff --git a/tools/virsh.pod b/tools/virsh.pod index 97d16c5..a6e9516 100644 --- a/tools/virsh.pod +++ b/tools/virsh.pod @@ -2958,6 +2958,24 @@ a node device between guest passthrough or the host. Libvirt will often do this action implicitly when required, but this command allows an explicit reset when needed. +=item B<nodedev-event> {[I<nodedev>] I<event> [I<--loop>] [I<--timeout> +I<seconds>] [I<--timestamp>] | I<--list>} + +Wait for a class of node device events to occur, and print appropriate +details of events as they happen. The events can optionally be filtered +by I<nodedev>. Using I<--list> as the only argument will provide a list +of possible I<event> values known by this client, although the connection +might not allow registering for all these events. + +By default, this command is one-shot, and returns success once an event +occurs; you can send SIGINT (usually via C<Ctrl-C>) to quit immediately. +If I<--timeout> is specified, the command gives up waiting for events +after I<seconds> have elapsed. With I<--loop>, the command prints all +events until a timeout or interrupt key. + +When I<--timestamp> is used, a human-readable timestamp will be printed +before the event. + =back =head1 VIRTUAL NETWORK COMMANDS -- 2.7.4

2 1

[libvirt] [PATCH] virt-admin: Properly fix the default session daemon URI to admin server
by Erik Skultety 09 Aug '16

09 Aug '16

Commit 30ce2f0e tried to fix the issue with an incorrect session URI to admin server but it messed up the checks: if (geteuid == 0 && VIR_STRDUP(*uristr, "libvirtd:///system") < 0) return -1; else if (VIR_STRDUP(*uristr, "libvirtd:///session") < 0) return -1; So if a client executed with root privileges tries to connect, its euid is checked (true) and the correct URI is successfully copied to @uristr (false), therefore the 'else' branch is taken and @uristr is replaced by the session URI which for root results in: Failed to connect socket to '/root/.cache/libvirt/libvirt-admin-sock': No such file or directory Signed-off-by: Erik Skultety <eskultet(a)redhat.com> --- src/libvirt-admin.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/src/libvirt-admin.c b/src/libvirt-admin.c index 4552e84..03245b6 100644 --- a/src/libvirt-admin.c +++ b/src/libvirt-admin.c @@ -179,11 +179,13 @@ virAdmGetDefaultURI(virConfPtr conf, char **uristr) * we set the default admin server URI to 'libvirtd:///system' or * 'libvirtd:///session' depending on the process's EUID. */ - if (geteuid() == 0 && - VIR_STRDUP(*uristr, "libvirtd:///system") < 0) + if (geteuid() == 0) { + if (VIR_STRDUP(*uristr, "libvirtd:///system") < 0) return -1; - else if (VIR_STRDUP(*uristr, "libvirtd:///session") < 0) + } else { + if (VIR_STRDUP(*uristr, "libvirtd:///session") < 0) return -1; + } } } -- 2.5.5

2 2

[libvirt] [PATCH] admin: Fix the default uri for session daemon to libvirtd:///session
by Erik Skultety 09 Aug '16

09 Aug '16

Just like we decide on which URI we go with based on EUID for qemu in remote driver, do a similar thing for admin except we do not spawn a daemon in this case. Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1356858 Signed-off-by: Erik Skultety <eskultet(a)redhat.com> --- src/libvirt-admin.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/src/libvirt-admin.c b/src/libvirt-admin.c index 4bf29b1..4552e84 100644 --- a/src/libvirt-admin.c +++ b/src/libvirt-admin.c @@ -176,10 +176,14 @@ virAdmGetDefaultURI(virConfPtr conf, char **uristr) /* Since we can't probe connecting via any hypervisor driver as libvirt * does, if no explicit URI was given and neither the environment * variable, nor the configuration parameter had previously been set, - * we set the default admin server URI to 'libvirtd://system'. + * we set the default admin server URI to 'libvirtd:///system' or + * 'libvirtd:///session' depending on the process's EUID. */ - if (VIR_STRDUP(*uristr, "libvirtd:///system") < 0) - return -1; + if (geteuid() == 0 && + VIR_STRDUP(*uristr, "libvirtd:///system") < 0) + return -1; + else if (VIR_STRDUP(*uristr, "libvirtd:///session") < 0) + return -1; } } -- 2.5.5

4 5

[libvirt] vfio: Error: Failed to setup INTx fd: Device or resource busy when PF pass-thought duel port NIC
by Moshe Levi 09 Aug '16

09 Aug '16

Hi, I am trying to PF pass-thought with Mellanox NIC CX4 with duel port card. I know the currently it doesn't support ACS, therefore we can't allocate 2 PF to 2 guest. (both pci device are in the same IOMMU group) But if I understand accoriding to [1] this currently there should be a problem to pass-thought both PFs to the same guest. before we boot the guest I unbind the pci device with the following commands. echo 0000:03:00.00> /sys/bus/pci/devices/0000:03:00.00/driver/unbind echo 0000:03:00.01 > /sys/bus/pci/devices/0000:03:00.1/driver/unbind I tried it worked and it was working on one environment , but on other environment we are getting the following error vfio: Error: Failed to setup INTx fd: Device or resource busy. See detail error [2]. Can anyone explain when/why this error occurs? And how can we fix it? [1] - http://vfio.blogspot.co.il/2014/08/iommu-groups-inside-and-out.html [2] - Nova-compute.log error: 2016-08-08 21:20:14.213 20236 ERROR nova.virt.libvirt.guest [req-5a6532d4-317f-4165-855e-37d8b36c71ec 295fe519b3864de5894fa0219362c26f 2e4cb36a0a3645e7b924deecea3bd210 - - -] Error launching a defined domain with XML: <domain type='kvm'> <name>instance-00000040</name> <uuid>74abfd25-aab6-4362-ab28-49c3bbd5e508</uuid> <metadata> <nova:instance xmlns:nova="http://openstack.org/xmlns/libvirt/nova/1.0"> <nova:package version="12.0.3-1.el7ost"/> <nova:name>Test</nova:name> <nova:creationTime>2016-08-08 21:20:08</nova:creationTime> <nova:flavor name="Mellanox_PT_2"> <nova:memory>16384</nova:memory> <nova:disk>0</nova:disk> <nova:swap>0</nova:swap> <nova:ephemeral>0</nova:ephemeral> <nova:vcpus>16</nova:vcpus> </nova:flavor> <nova:owner> <nova:user uuid="295fe519b3864de5894fa0219362c26f">admin</nova:user> <nova:project uuid="2e4cb36a0a3645e7b924deecea3bd210">openstack</nova:project> </nova:owner> <nova:root type="image" uuid="49d20d6e-e247-43ff-aeec-84ba53d6b439"/> </nova:instance> </metadata> <memory unit='KiB'>16777216</memory> <currentMemory unit='KiB'>16777216</currentMemory> <vcpu placement='static' cpuset='3-23,27-47'>16</vcpu> <cputune> <shares>16384</shares> </cputune> <sysinfo type='smbios'> <system> <entry name='manufacturer'>Red Hat</entry> <entry name='product'>OpenStack Compute</entry> <entry name='version'>12.0.3-1.el7ost</entry> <entry name='serial'>d336b2a1-4abd-4602-9bc4-4aeb20f778e7</entry> <entry name='uuid'>74abfd25-aab6-4362-ab28-49c3bbd5e508</entry> <entry name='family'>Virtual Machine</entry> </system> </sysinfo> <os> <type arch='x86_64' machine='pc-i440fx-rhel7.2.0'>hvm</type> <boot dev='hd'/> <smbios mode='sysinfo'/> </os> <features> <acpi/> <apic/> </features> <cpu mode='host-model'> <model fallback='allow'/> <topology sockets='16' cores='1' threads='1'/> </cpu> <clock offset='utc'> <timer name='pit' tickpolicy='delay'/> <timer name='rtc' tickpolicy='catchup'/> <timer name='hpet' present='no'/> </clock> <on_poweroff>destroy</on_poweroff> <on_reboot>restart</on_reboot> <on_crash>destroy</on_crash> <devices> <emulator>/usr/libexec/qemu-kvm</emulator> <disk type='file' device='disk'> <driver name='qemu' type='qcow2' cache='none'/> <source file='/var/lib/nova/instances/74abfd25-aab6-4362-ab28-49c3bbd5e508/disk'/> <target dev='vda' bus='virtio'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x04' function='0x0'/> </disk> <controller type='usb' index='0'> <address type='pci' domain='0x0000' bus='0x00' slot='0x01' function='0x2'/> </controller> <controller type='pci' index='0' model='pci-root'/> <interface type='bridge'> <mac address='fa:16:3e:14:ec:96'/> <source bridge='qbre5243bb0-8e'/> <target dev='tape5243bb0-8e'/> <model type='virtio'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x03' function='0x0'/> </interface> <serial type='file'> <source path='/var/lib/nova/instances/74abfd25-aab6-4362-ab28-49c3bbd5e508/console.log'/> <target port='0'/> </serial> <serial type='pty'> <target port='1'/> </serial> <console type='file'> <source path='/var/lib/nova/instances/74abfd25-aab6-4362-ab28-49c3bbd5e508/console.log'/> <target type='serial' port='0'/> </console> <input type='tablet' bus='usb'/> <input type='mouse' bus='ps2'/> <input type='keyboard' bus='ps2'/> <graphics type='vnc' port='-1' autoport='yes' listen='0.0.0.0' keymap='en-us'> <listen type='address' address='0.0.0.0'/> </graphics> <video> <model type='cirrus' vram='16384' heads='1'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x02' function='0x0'/> </video> <hostdev mode='subsystem' type='pci' managed='yes'> <source> <address domain='0x0000' bus='0x03' slot='0x00' function='0x1'/> </source> <address type='pci' domain='0x0000' bus='0x00' slot='0x05' function='0x0'/> </hostdev> <hostdev mode='subsystem' type='pci' managed='yes'> <source> <address domain='0x0000' bus='0x03' slot='0x00' function='0x0'/> </source> <address type='pci' domain='0x0000' bus='0x00' slot='0x06' function='0x0'/> </hostdev> <memballoon model='virtio'> <stats period='10'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x07' function='0x0'/> </memballoon> </devices> </domain> 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [req-5a6532d4-317f-4165-855e-37d8b36c71ec 295fe519b3864de5894fa0219362c26f 2e4cb36a0a3645e7b924deecea3bd210 - - -] [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] Instance failed to spawn 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] Traceback (most recent call last): 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] File "/usr/lib/python2.7/site-packages/nova/compute/manager.py", line 2156, in _build_resources 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] yield resources 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] File "/usr/lib/python2.7/site-packages/nova/compute/manager.py", line 2009, in _build_and_run_instance 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] block_device_info=block_device_info) 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] File "/usr/lib/python2.7/site-packages/nova/virt/libvirt/driver.py", line 2585, in spawn 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] block_device_info=block_device_info) 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] File "/usr/lib/python2.7/site-packages/nova/virt/libvirt/driver.py", line 4699, in _create_domain_and_network 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] xml, pause=pause, power_on=power_on) 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] File "/usr/lib/python2.7/site-packages/nova/virt/libvirt/driver.py", line 4629, in _create_domain 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] guest.launch(pause=pause) 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] File "/usr/lib/python2.7/site-packages/nova/virt/libvirt/guest.py", line 142, in launch 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] self._encoded_xml, errors='ignore') 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] File "/usr/lib/python2.7/site-packages/oslo_utils/excutils.py", line 204, in __exit__ 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] six.reraise(self.type_, self.value, self.tb) 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] File "/usr/lib/python2.7/site-packages/nova/virt/libvirt/guest.py", line 137, in launch 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] return self._domain.createWithFlags(flags) 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] File "/usr/lib/python2.7/site-packages/eventlet/tpool.py", line 183, in doit 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] result = proxy_call(self._autowrap, f, *args, **kwargs) 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] File "/usr/lib/python2.7/site-packages/eventlet/tpool.py", line 141, in proxy_call 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] rv = execute(f, *args, **kwargs) 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] File "/usr/lib/python2.7/site-packages/eventlet/tpool.py", line 122, in execute 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] six.reraise(c, e, tb) 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] File "/usr/lib/python2.7/site-packages/eventlet/tpool.py", line 80, in tworker 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] rv = meth(*args, **kwargs) 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] File "/usr/lib64/python2.7/site-packages/libvirt.py", line 1059, in createWithFlags 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] if ret == -1: raise libvirtError ('virDomainCreateWithFlags() failed', dom=self) 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] libvirtError: internal error: early end of file from monitor: possible problem: 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] 2016-08-08T21:20:13.577395Z qemu-kvm: -device vfio-pci,host=03:00.0,id=hostdev1,bus=pci.0,addr=0x6: vfio: Error: Failed to setup INTx fd: Device or resource busy 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] 2016-08-08T21:20:13.578216Z qemu-kvm: -device vfio-pci,host=03:00.0,id=hostdev1,bus=pci.0,addr=0x6: Device initialization failed 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] 2016-08-08T21:20:13.578231Z qemu-kvm: -device vfio-pci,host=03:00.0,id=hostdev1,bus=pci.0,addr=0x6: Device 'vfio-pci' could not be initialized 2016-08-08 21:20:14.214 20236 ERROR nova.compute.manager [instance: 74abfd25-aab6-4362-ab28-49c3bbd5e508] 2016-08-08 21:20:54.118 20236 WARNING nova.compute.manager [req-d1bd757e-ab71-472b-8555-935bfbecbe89 - - - - -] Bandwidth usage not supported by hypervisor.

1 0

[libvirt] ANNOUNCE: libguestfs 1.34 released
by Richard W.M. Jones 09 Aug '16

09 Aug '16

I'm pleased to announce libguestfs 1.34, a library and set of tools for accessing and modifying virtual machine disk images. This release took about 8 months of work by many contributors. Amongst the new features are large performance improvements, substantial enhancements to virt-p2v & virt-v2v, better SELinux support, and APIs for doing forensic analysis of disk images. See the release notes below for full details. You can get libguestfs 1.34 here: Main website: http://libguestfs.org/ Source: http://libguestfs.org/download/1.34-stable/ Fedora 24: http://koji.fedoraproject.org/koji/packageinfo?packageID=8391 It will appear as an update for F24 in about a week. Fedora 25: (blocked on RHBZ#1365270) Debian/experimental: https://packages.debian.org/libguestfs0 Rich. Release notes (also available online at http://libguestfs.org/guestfs-release-notes.1.html ): New features Multiple performance enhancements were made in libguestfs. The "hot cache" time to launch to appliance should be under 1 second assuming recent qemu and kernel are installed. There are also new utilities for precisely benchmarking libguestfs (utils/boot-benchmark and utils/boot-analysis in the source tree). The virt-p2v tool for converting physical machines to virtual machines was substantially improved. This includes: clearer, coloured output during conversions, support for Gtk 3, more detailed information about hardware, click to identify network interfaces, more debugging tools included with the ISO, and many bug fixes. f2fs (Flash Friendly File System) is now supported (Pino Toscano). New tools virt-p2v-make-kiwi(1) can be used to build the virt-p2v ISO based on SLES and openSUSE, using the kiwi utility (Cédric Bosdonnat). New features in existing tools virt-resize will now preserve the GPT GUID. This was required for Windows Server 2012 R2, where the bootloader would become confused if the GUID changed (Maxim Perevedentsev). virt-resize will use sparse copying for (old MBR-style) extended partitions. This makes resizing of guests that use extended partitions much faster (Maxim Perevedentsev). virt-p2v kernel command line options can now be used to set defaults for GUI configuration. The virt-p2v debugging options have been completely removed, simplifying the interface and documentation. Debugging information is now captured fully automatically. virt-p2v-make-disk lets you specify an --arch option, allowing you to build a 32 bit virt-p2v, for compatibility with older systems. virt-p2v-make-disk no longer requires that you specify an "os-version" for the virt-p2v disk. If omitted it will try to choose a suitable "os-version" depending on your host system. virt-p2v-make-disk and virt-p2v-make-kickstart both gain a new --install option that allows you to add arbitrary extra packages to the virt-p2v ISO, for customization, additional debugging tools and so on. virt-v2v will now uninstall Parallels Tools (or the equivalent Virtuozzo Tools) from Linux guests. Also stop the Windows drivers from loading at boot. (Roman Kagan and Pavel Butsykin) virt-v2v --in-place mode has been enhanced to allow the caller to choose whether or not to install certain virtio drivers in the guest (Roman Kagan). virt-v2v conversion of Windows guests was substantially rewritten and simplified (Roman Kagan). virt-v2v --in-place mode now supports installing virtio-scsi drivers in guests (Roman Kagan). virt-v2v can now convert SUSE guests and SUSE guests using UEFI (Cédric Bosdonnat and Jim Fehlig). virt-v2v can now convert guests to Glance that have multiple disks. Previously it would fail on such guests. The virt-v2v --no-trim and --vmtype options are now no-ops. They will print a warning but are otherwise ignored. virt-v2v can now generate the OVF vmtype correctly without user intervention. virt-v2v has now been tested against Citrix Xen as a source hypervisor (Cédric Bosdonnat). virt-v2v adds support for SUSE VMDP drivers (Cédric Bosdonnat). virt-v2v can convert OVA files containing subfolders, as produced by SUSE Studio (Cédric Bosdonnat). virt-v2v sets the OVF "<Origin>" element correctly. oVirt has been extended to support more source hypervisors (Shahar Havivi). virt-v2v now supports Windows Server 2016 (Tomáš Golembiovský). The virt-builder --list option can now be used to show all templates or a single template (Pino Toscano). All OCaml-based tools now use getopt_long(3) for option parsing, and --help output has been improved (Pino Toscano). virt-builder and virt-customize --selinux-relabel option can now fully relabel the guest filesystem at build time, without requiring a lengthy autorelabel at first boot. virt-customize --delete now accepts globs. New virt-customize --uninstall option lets you uninstall packages. virt-customize can now use "pvvxsvc" as an alternative to "rhsrvany" for running firstboot scripts in Windows guests (Cédric Bosdonnat). virt-customize now uses the strongest hashing scheme for passwords on Arch and Void Linux (Pino Toscano). virt-customize --install now works correctly on Arch (Pino Toscano). virt-inspector has new options --no-applications and --no-icon to prevent the list of applications and icon from being included in the XML output (Pino Toscano). New virt-sysprep --network option has been added, allowing you to actually use the --install etc options which were present in virt- sysprep before but did not usually work. Note that the network is still disabled by default. virt-sysprep "fs-uuids" operation no longer fails on btrfs guests (Maxim Perevedentsev). virt-dib can output Docker images (Pino Toscano). virt-dib has a new --drive-format option to allow the user to specify the format of the helper drive (Pino Toscano). All OCaml virt tools now have a --colors/--colours option which enables coloured output (using ANSI escape sequences) even if the output is not a tty. The default is to check if the output is a tty and disable coloured output if not. This allows coloured output to be consumed by other tools. Language bindings PHP test coverage has been enhanced (Pino Toscano). PHP 7 is now supported (Pino Toscano). Python bindings are now compliant with PEP 8 (Pino Toscano). A Python pip package is available in http://libguestfs.org/download/python/ The Ruby bindings now print the full exception if one is thrown by the event callback. Note this is still incorrect behaviour as event callbacks should not throw exceptions, but it aids debugging. All OCaml libraries and programs are now compiled with -safe-string, if supported by the OCaml compiler. Inspection Alpine Linux using busybox can now be inspected. Also the APK package manager is supported in virt-customize (Pino Toscano). We now handle inspection of Mageia 4 (Pino Toscano). Void Linux and the Void Linux xbps package manager are fully supported (Pino Toscano). Parsing of CoreOS version information has been enhanced (Pino Toscano). It is now possible to get an icon from ALT Linux (Pino Toscano). PLD Linux versions < 3 are now recognized (Pino Toscano). Windows drive letters are now returned for guests using GPT partitions (Dawid Zamirski). We can now correctly inspect Unix guests that do not have an /etc/fstab file (Pino Toscano). Added another source for the Ubuntu icon which doesn't rely on GNOME having been installed in the guest. We can now get an icon for Windows 7 64 bit guests. Libosinfo integration was rewritten to deal with the new database format used by osinfo (Pino Toscano). Documentation New manual page guestfs-building(1) describes how to build libguestfs from source. The man pages, tools and tool --help output is now automatically checked to ensure that all tool options are properly documented, that warning sections are included where necessary, and that every page has a description section. The guestfs-testing(1) man page has been refreshed and based on a newer libguestfs. Architectures and platforms virt-customize now works on POWER7 and POWER8 platforms, both big endian and little endian (Xianghua Chen and Hu Zhang). Security See also guestfs-security(1). CVE-2015-8869 https://bugzilla.redhat.com/CVE-2015-8869 This vulnerability in OCaml might affect virt tools written in the OCaml programming language. It affects only 64 bit platforms. Because this bug affects code generation it is difficult to predict which precise software could be affected, and therefore our recommendation is that you recompile libguestfs using a version of the OCaml compiler where this bug has been fixed (or ask your Linux distro to do the same). virt-customize ownership of .ssh, .ssh/authorized_keys https://bugzilla.redhat.com/1337561 Previously when virt-customize injected an SSH key into a guest, when it created the ~/.ssh and ~/.ssh/authorized_keys directory and file (in case they were missing) it created them with owner and group "root.root". This has been fixed so the correct user is used. This is not thought to have been exploitable. Windows "%systemroot%" The inspection code has been made more robust against guests which might use very long "%systemroot%" (derived from the guest-controlled Windows Registry). This is not thought to have been exploitable. Virtio-rng is now available in the appliance virtio-rng (the virtual Random Number Generator device) is now passed to the appliance, which should improve the quality random numbers generated for GUIDs and cryptographic key generation. API New APIs "btrfs_filesystem_show" List all devices where a btrfs filesystem is spanned (Pino Toscano). "download_blocks" "download_inode" "filesystem_walk" Download filesystem data blocks from a given partition. Download arbitrary files by inode number. Retrieve all files from a filesystem including deleted files. Note these require optional dependency The Sleuth Kit. (Matteo Cafasso) "get_sockdir" Read the path where temporary sockets are stored (Pino Toscano). "mountable_device" "mountable_subvolume" Split a Mountable into device name and subvolume (Cédric Bosdonnat). "ntfscat_i" Download NTFS file by inode number (Matteo Cafasso). "part_expand_gpt" Allow in-place expanding of GPT partitions by moving the second (backup) partition table to the end of the disk (Maxim Perevedentsev). "part_get_disk_guid" "part_set_disk_guid" "part_set_disk_guid_random" Get and set the GPT disk GUID, or set it to a fresh random value (Maxim Perevedentsev). "selinux_relabel" SELinux-relabel part or all of the guest filesystem. Other API changes "guestfs_set_selinux", "guestfs_get_selinux", "guestfs_setcon", "guestfs_getcon" and "guestfs_llz" have been deprecated. Use the new API "guestfs_selinux_relabel" to relabel filesystems. Use "guestfs_lgetxattrs" to list the "security.selinux" extended attributes of existing files. "guestfs_vfs_minimum_size" can now be used on dirty filesystems (Maxim Perevedentsev). "guestfs_ll" now works on paths which contain absolute symlinks (Pino Toscano). "guestfs_glob_expand" now has an optional "directoryslash" boolean parameter which controls whether trailing slashes are returned for directory names (Pino Toscano). "guestfs_lvs" will no longer return LVs which have the "activationskip" flag set. The reason is that such LVs have no "/dev/VG/LV" device node and so code which read the list of LVs and then probed the devices themselves would immediately fail. You can use "guestfs_lvs_full" if you want to read all LVs. (Pino Toscano). "guestfs_list_disk_labels" now no longer fails if no disks with labels were added. Instead it now returns an empty list (Pino Toscano). "guestfs_is_lv" no longer fails if passed a btrfs subvolume, it returns false instead (Maxim Perevedentsev). Build changes qemu ≥ 1.3.0 is required. yajl (a JSON parsing library) is required to build libguestfs. You can now build with GCC 6. "make check-valgrind" now has substantially better coverage. "make check-slow" now works again. Use "make -C appliance clean-supermin-appliance" to clean the supermin appliance (it will be rebuilt on next "make"). There are a variety of new rules for running virt-p2v from the source directory: "make -C p2v run-virt-p2v-directly" | "run-virt-p2v-in-a-vm" | "run-virt-p2v-non-gui-conversion". These are documented further in guestfs-hacking(1). virt-p2v may be built using either Gtk 2 or Gtk 3. To force a particular version of Gtk to be used, "./configure --with-gtk=2|3" The "./configure" options are now mostly documented in guestfs-building(1). Internals In git, versions are now tagged with "v1.XX.YY" (previously they were tagged with "1.XX.YY"). Using the "v-" prefix is more common in git repositories. When using the libvirt backend, we now wait for qemu to exit gracefully instead of killing it after 15 seconds. This helps when writing to slow devices (especially cheap USB keys). Error messages from libvirt now include the "err->int1" field which usually contains the "errno". On ARM, all DTB (device tree) code has been removed. qemu creates the right device tree on the fly, we do not need to specify one. The C API tests now use larger test disks, allowing BTRFS to be tested properly (Pino Toscano). The tests should now work on a pure Python 3 host (Pino Toscano). In C bindings, internal functions are now (mostly) consistently named "guestfs_int_*" whereas previously there was no consistent scheme. The old "safe_malloc" etc functions are now no longer exported by the library, nor used in language bindings. Setting TMPDIR to a path longer than ~ 100 characters will no longer cause libguestfs to fail silently and randomly when creating Unix domain sockets (Pino Toscano). The "COMPILE_REGEXP" macro can now be used in the daemon. When tracing, results containing structs are now printed in full (Pino Toscano). The Perl "Sys::Guestfs" module now no longer embeds an incrementing API "version number". This module is now always at phony version "1.0". To find the real version of libguestfs from Perl you must call "$g->version". All code is compiled with "-Wstack-usage=10000" and multiple changes have been made to remove stack allocation of large strings and buffers. The error(3) function is now used everywhere, replacing most previous uses of perror(3) + exit(3), and fprintf(3) + exit. In C code, "/**" comments are turned into documentation which is automatically added to the guestfs-hacking(1) manual page. A safe "getumask" function has been added. For recent Linux kernels this uses the newly added "Umask" field in /proc/self/status. For older Linux and other Unix, this uses a thread-safe technique involving fork(2) (thanks: Josh Stone, Jiri Jaburek, Eric Blake). Safe posix_fadvise(2) wrappers have been added, and more hints have been added to the code which may make a minor difference to performance. A safe wrapper around waitpid(2) has been added which handles "INTR" properly. "podwrapper.pl" (used to generate the manual pages) now stops if any POD error is found. A new script called "podcheck.pl" does cross- checking of --help output, tool options and manual pages. All version numbers in the library (eg. versions of qemu, versions of libvirt, versions of guest operating systems) are unified in a single file src/version.c (Pino Toscano). On Windows guests, virt-customize will use the vendor-neutral path "C:\Program Files\Guestfs\Firstboot" to store firstboot scripts. Previously it used "C:\Program Files\Red Hat\Firstboot". This change should be invisible to the scripts themselves. (Cédric Bosdonnat) On Linux guests, the firstboot services generated by virt-builder --firstboot etc have been renamed to "guestfs-firstboot" (Pino Toscano). There is now a common "debug" function used by all OCaml tools, replacing previous code which did "if verbose () then printf ...". virt-p2v copies files it needs over to the virt-v2v conversion server using scp(1), instead of trying to send them via the shell session. This should improve reliability and should be a completely transparent to end users. All code in mllib is now built into a single "mllib.cma" or "mllib.cmxa" library. All code in customize is now built into a single "customize.cma" or "customize.cmxa" library. This simplifies the build of the OCaml tools. lvmetad(8) is now used in the appliance when available (Pino Toscano). "Silent rules" are used for OCaml, Java, Erlang and POD. Use "make V=1" to see the full command lines again (Pino Toscano). Bugs fixed https://bugzilla.redhat.com/1364347 virt-sparsify --in-place failed with UEFI system https://bugzilla.redhat.com/1362357 run_command runs exit handlers when execve fails (e.g. due to missing executable) https://bugzilla.redhat.com/1362354 virt-dib failed to create image using DIB_YUM_REPO_CONF https://bugzilla.redhat.com/1359652 Fail to inspect Windows ISO file https://bugzilla.redhat.com/1358142 Some info will show when convert guest to libvirt by virt-v2v with parameter --quiet https://bugzilla.redhat.com/1354335 overlay of disk images does not specify the format of the backing file https://bugzilla.redhat.com/1352761 Virt-manager can't show OS icons of win7/win8/ubuntu guest. https://bugzilla.redhat.com/1350363 Improve error info "remote server timeout unexpectedly waiting for password prompt" when connect to a bogus server at p2v client https://bugzilla.redhat.com/1348900 virt-p2v should update error prompt when 'Test connection' with a non-existing user in conversion server https://bugzilla.redhat.com/1345813 virt-sysprep --install always failed to install the packages specified https://bugzilla.redhat.com/1345809 virt-customize --truncate-recursive should give an error message when specifying a no-existing path https://bugzilla.redhat.com/1343423 [RFE]Should give a better description about 'curl error 22' when failed using ssh identity http url at p2v client https://bugzilla.redhat.com/1343414 Failed SSH to conversion server by ssh identity http url at p2v client https://bugzilla.redhat.com/1343375 [RFE] uninstall packages inside the VM https://bugzilla.redhat.com/1342447 Ifconfig command is not supported on p2v client https://bugzilla.redhat.com/1342398 Convert a guest from RHEL by virt-v2v but its origin info shows RHEV at rhevm https://bugzilla.redhat.com/1342337 Should remind a warning about disk image has a partition when using virt-p2v-make-disk https://bugzilla.redhat.com/1341984 virt-get-kernel prompts an 'invalid value' error when using --format auto https://bugzilla.redhat.com/1341564 virt-p2v spinner should be hidden when it stops spinning https://bugzilla.redhat.com/1340809 Testing connection timeout when input regular user of conversion server with checked "use sudo......"button https://bugzilla.redhat.com/1340464 [RFE] Suggestion give user a reminder for "Cancel conversion" button https://bugzilla.redhat.com/1340407 Multiple network ports will not be aligned at p2v client https://bugzilla.redhat.com/1338083 Update UEFI whitelist for official fedora packages https://bugzilla.redhat.com/1337561 virt-customize --ssh-inject not applying correct file permission https://bugzilla.redhat.com/1335671 extra quotes around UUID confuses findfs in RHEL (but not in Fedora) https://bugzilla.redhat.com/1332025 Inspection does not parse /etc/redhat-release containing "Derived from Red Hat Enterprise Linux 7.1 (Source)" https://bugzilla.redhat.com/1327488 RFE: Allow p2v kernel options without p2v.server to set defaults https://bugzilla.redhat.com/1325825 virt-v2v should prevent using multiple '-b' and '-n' option appears on the command line https://bugzilla.redhat.com/1321620 libguestfs: error: could not parse integer in version number: 7" https://bugzilla.redhat.com/1321338 [1.33.16] Compilation Error: Unbound value List.sort_uniq in v2v.ml line 988, characters 10-24: https://bugzilla.redhat.com/1317843 `virt-builder --update` fails with: "dnf -y --best upgrade: command exited with an error" https://bugzilla.redhat.com/1316479 v2v cmd cannot exit and "block I/O error in device 'appliance': No space left on device (28)" is printed when specified "-v -x" https://bugzilla.redhat.com/1316041 virt-rescue fails, but missing error message https://bugzilla.redhat.com/1314244 RFE: virt-p2v log window should process colour escapes and backspaces https://bugzilla.redhat.com/1312254 virt-v2v -o libvirt doesn't preserve or use correct <graphics type="vnc|spice"> https://bugzilla.redhat.com/1309706 error: internal error: Invalid floppy device name: hdb https://bugzilla.redhat.com/1309619 Wrong warning info "use standard VGA" shows when converting windows > 7 by virt-v2v https://bugzilla.redhat.com/1309580 OS name of win8.1 x64 guest shows incorrect in rhevm3.6 general info https://bugzilla.redhat.com/1308769 virt-v2v does not copy additional disks to Glance https://bugzilla.redhat.com/1306666 Failure when disk contains an LV with activationskip=y https://bugzilla.redhat.com/1296606 virt-v2v doesn't remove VirtualBox additions correctly because of file quoting https://bugzilla.redhat.com/1293527 There should be a reminder to avoid user to edit a guest image by multiple tools at the same time in guestfish man page https://bugzilla.redhat.com/1293276 guestfish can not ll a symbolic link dir or edit a file in it https://bugzilla.redhat.com/1278878 guestfish should be able to handle LVM thin layouts https://bugzilla.redhat.com/1264835 ppc64le: virt-customize --install fail to detect the guest arch https://bugzilla.redhat.com/1264332 Test that trimming in virt-v2v doesn't regress https://bugzilla.redhat.com/1232192 Virt-v2v gives an error on a blank disk: part_get_parttype: unknown signature, of the output: BYT; https://bugzilla.redhat.com/1229386 virt-p2v in non-GUI mode doesn't show any conversion progress or status https://bugzilla.redhat.com/1227599 P2V invalid password prints unexpected end of file waiting for command prompt. https://bugzilla.redhat.com/1224795 On Ubuntu, virt-builder --install and --update cannot use the network https://bugzilla.redhat.com/1213324 virt-v2v: warning: unknown guest operating system: windows windows 6.3 when converting win8,win8.1,win2012,win2012R2,win10 to rhev https://bugzilla.redhat.com/1203898 Support inspecting docker images without /etc/fstab https://bugzilla.redhat.com/1186935 libguestfs cannot inspect recent Fedora / RHEL >= 7 when /usr is a separate partition https://bugzilla.redhat.com/1167916 P2V: invalid conversion server prints unexpected end of file waiting for password prompt. https://bugzilla.redhat.com/1152825 virt-rescue --selinux can not work well, when enable selinux in the command line the value of 'getenforce' is still Disabled in virt- rescue appliance https://bugzilla.redhat.com/1150298 ARM 32 bit on Ubuntu: warning: cast to pointer from integer of different size [-Wint-to-pointer-cast] https://bugzilla.redhat.com/1089100 NetworkManager avc unlink denied for resolv.conf after using --selinux-relabel https://bugzilla.redhat.com/983969 RFE: virt-sysprep should be SELinux-aware https://bugzilla.redhat.com/855058 RFE: virt-p2v: display more information about storage devices https://bugzilla.redhat.com/554829 SELinux handling could be done better. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-p2v converts physical machines to virtual machines. Boot with a live CD or over the network (PXE) and turn machines into KVM guests. http://libguestfs.org/virt-v2v

1 0

[libvirt] Virtqueue size exceeded error when resuming VM
by Moshe Levi 08 Aug '16

08 Aug '16

Hi, A new security fix [1],[2] and [3] merged to qemu. After updating the packages we started to get "qemu-system-x86_64: Virtqueue size exceeded", when resuming the guest. Our environment is OpenStack master and we have Mellanox CI that test SR-IOV functionality. Ubuntu 14.04 with Qemu 2.0.0+dfsg-2ubuntu1.26 that contains the fixes see [2] ii qemu-kvm 2.0.0+dfsg-2ubuntu1.26 amd64 QEMU Full virtualization ii qemu-system-x86 2.0.0+dfsg-2ubuntu1.26 amd64 QEMU full system emulation binaries (x86) ii qemu-utils 2.0.0+dfsg-2ubuntu1.26 amd64 QEMU utilities Our CI started to fail last week when this security packages released. The scenarios is as follows (sorry for the OpenStack commands :)) : 1. nova boot guest 2. nova suspend guest 3. nova resume guest The result is that the guest is in poweroff state and when I power it on everything is working fine. I tested in direct port (SR-IOV) and normal port (virtual port) and it happens in both cases. According to the [3] it prevent from malicious guest to submit more requests than the virtqueuesize permits. Our CI uses proprietary Cirros image with mlnx4_en driver. (http://13.69.151.247/images/mellanox_eth.img) I started to test it with other images to see if the problem with our image. I also tested with Ubuntu image - https://cloud-images.ubuntu.com/wily/current/wily-server-cloudimg-amd64-dis… And OpenStack Cirros image http://download.cirros-cloud.net/0.3.4/cirros-0.3.4-x86_64-disk.img The Ubuntu image had the same failure, but the Cirros worked. I wonder if there is a problem with the patch or with the images? What in these images can make them malicious guest? [1] - https://access.redhat.com/security/cve/cve-2016-5403 [2] - http://www.ubuntu.com/usn/usn-3047-1/ [3] - https://lists.gnu.org/archive/html/qemu-devel/2016-07/msg06257.html

1 0