- Users - libvirt List Archives

[libvirt-users] is memoryBacking support 'share' and 'mem-path' parameter
by Linhaifeng 09 Oct '14

09 Oct '14

Hi, I want to use this qemu command '-object memory-backend-file,id=mem,size=2048M,mem-path=/mnt/huge,share=on' with libvirt but i can't find the 'mem-path' and 'share' in the documentation. because the vhost-user backend based on 'share=on' parameter and libvirt support vhostuser i guess there maybe another way to support this parameter? this is my xml: <cpu> <numa> <cell id='0' cpus='0' memory='1048576' memAccess='shared'/> <cell id='1' cpus='1' memory='1048576' memAccess='shared'/> </numa> </cpu> <memoryBacking> <hugepages> <page size="2" unit="M" nodeset="0,1"/> </hugepages>   </memoryBacking> <interface type='vhostuser'> <mac address='52:54:00:3b:83:1a'/> <source type='unix' path='/path/port1' mode='client'/> <model type='virtio'/> </interface> <interface type='vhostuser'> <mac address='52:54:00:3b:83:1b'/> <source type='unix' path='/path/port2' mode='client'/> <model type='virtio'/> </interface> when i create the VM the qemu command is like this without 'share=on' parameter: -object memory-backend-file,prealloc=yes,mem-path=/dev/hugepages/libvirt/qemu,size=1024M,id=ram-node0 -numa node,nodeid=0,cpus=0,memdev=ram-node0 -object memory-backend-file,prealloc=yes,mem-path=/dev/hugepages/libvirt/qemu,size=1024M,id=ram-node1 -numa node,nodeid=1,cpus=1,memdev=ram-node1

2 3

[libvirt-users] Question about vnc connection password set to empty
by Wei Zhang 09 Oct '14

09 Oct '14

Hello, everybody here I have a question about graphical buffers for rhel7. Thanks so much in advance who can give me some suggestions about that. When I try to do regression for bug859317(https://bugzilla.redhat.com/show_bug.cgi?id=859317) on RHEL7， I found that guest could be connected success with virt-viewer when set/update the graphics passwd to empty. But according to comment 2 in this bug, libvirt will blocks all access to the VNC server when set a password to empty. So could you please help us check is it acceptable on RHEL7? Below is my test steps on RHEL7: Version-Release number of selected component (if applicable): libvirt-1.2.8-3.el7.x86_64 qemu-kvm-rhev-2.1.0-5.el7.x86_64 kernel-3.10.0-150.el7.x86_64 How reproducible: 100% Steps to Reproduce: 1. Define guest with a graphics device like: ......................................... <graphics type='vnc' port='-1' autoport='yes' passwd=""/> ......................................... 2. Connecting with vncviewer or virt-viewer . # virsh start dom2 # virsh dumpxml dom2 --security-info |grep graphics <graphics type='vnc' port='5900' autoport='yes' listen='127.0.0.1' passwd=''> </graphics> # virt-viewer dom2 connect the guest success with empty password 3. The problem can be reproduced when changed the graphics passwd from an non-empty to an empty # virsh dumpxml dom2 --security-info |grep graphics <graphics type='vnc' port='5900' autoport='yes' listen='127.0.0.1' passwd='123'> </graphics> # cat change-passwd.xml <graphics type='vnc' port='-1' autoport='yes' listen='127.0.0.1' passwd=''/> # virsh update-device dom2 change-passwd.xml # virsh dumpxml dom2 --security-info |grep graphics <graphics type='vnc' port='5900' autoport='yes' listen='127.0.0.1' passwd=''> </graphics> login the guest with virt-viewer command # virt-viewer testks also can connect the guest dom2 success with empty password 4. I produce the same steps on RHEL6.6 libvirt-0.10.2-46.el6.x86_64, the guest can not be connected when set graphics vnc password to empty, the same result with bug 859317. Actual results: we can login the guest when set the graphics passwd empty Expected results: I am not sure what is the right behaviour for RHEL7? thank

1 0

[libvirt-users] virsh net-start default fails
by David Roundy 08 Oct '14

08 Oct '14

Hi all, I'm getting an error, and have failed to succeed in googling for help with it. I have an old VM that I created a while back (at least one reboot ago, possibly a kernel upgrade ago), and am now finding that it won't start due to a lack of the 'default' network. When I look for how to fix this, everyone seems to agree on the command: # virsh net-start default error: failed to get network 'default' error: An error occurred, but the cause is unknown But this fails. A natural next check would be to use net-list: # virsh net-list --all Name State Autostart ----------------------------------------- Which demonstrates that the 'default' network is unknown. I'm certain that it used to exist, and the file /var/lib/libvirt/network/default.xml does exist, claims to be autogenerated and looks reasonable. I am at a loss as to how to debug this further. Any suggestions? -- David Roundy

4 3

[libvirt-users] sr-vio on intel while virsh chooses rtl8139 for model type
by lejeczek 08 Oct '14

08 Oct '14

hi everybody I'd presume virsh makes the best possible choice, right? It is that just seems bit... odd having realtek in guest and Intel's VF on host, no? regards

2 5

[libvirt-users] win2k8 guest and (strange) number of cpus task manager sees
by lejeczek 08 Oct '14

08 Oct '14

hi everybody a qemu-kvm guest gets 16 cpus and Windows in "Device Manager" sees all sixteeen but "Task Manager" shows only 4, the same "System" properties say "(4 processors)" I'd like to learn a bit about this - is it some sort of "resources management" on the libvirt/qemu causing this? And what does it do to a guest performance-wise side? it's on Scientific L 7, qemu-kvm-1.5.3-60.el7_0.7.x86_64 libvirt-1.1.1-29.el7_0.1.x86_64 regards P.

2 2

[libvirt-users] what is the xml fomat about memory-backend-file
by Linhaifeng 08 Oct '14

08 Oct '14

Hi, I want to use virsh to create a VM with the qemu parameter '-object memory-backend-file,id=mem,size=2048M,mem-path=/mnt/huge,share=on -numa node,memdev=mem'. How to write the XML file?

3 4

[libvirt-users] libvirt-sandbox on Fedora 21
by Gene Czarcinski 07 Oct '14

07 Oct '14

It sure would be nice if someone who knows what they are doing took a look at this problem: https://bugzilla.redhat.com/show_bug.cgi?id=1135762 The current situation on Fedora 21 is that virt-sandbox and virt-sandbox-service simply do not work if a network is defined (e.g., -N address-192.168.1.1/34). I would give dhcp as an example but to work, that needs to patch I submitted in another BZ report. Since both virt-sandbox -c lxc:/// -N <network>,source=default and doing systemctl start <sandbox> result in the same error and since I am running the fedora-virt-preview rpms and the failure *does not* occur on Fedora 20 but *does* occur on Fedora 21, I have to assume that the problem is not directly in libvirt. However, determining what/where the problem really is, is stretching my abilities a bit too far ... help!! Gene

2 3

[libvirt-users] Virtual gaming/workstation desktops using KVM VGA passthrough?
by Alex G.S. 07 Oct '14

07 Oct '14

Dear List, Recently I read an article [1] about how to use KVM and VGA passthrough to create multiple GPU accelerated VM's for use in gaming or as virtual workstations: [1] http://www.pugetsystems.com/labs/articles/Multiheaded-NVIDIA-Gaming-using-U… I'm planning on building a KVM workstation to do this using CentOS 7 or Ubuntu 14.04 LTS as the base virtualization host and then an Intel Xeon CPU and two Nvidia GeForce GTX 750's as GPU's for the virtual machines. I'm really new to this sort of GPU passthrough on KVM and am wondering if anyone has any advice or has done this before? Thank you!

3 4

[libvirt-users] passthrough of PCI-device
by Weis, Michael (DWIE) 07 Oct '14

07 Oct '14

Hello, I try to passthrough a PCI-card to a VM named testvm I want to do that with an xml-file named hga.xml including the following content: <hostdev mode='subsystem' type='pci' managed='yes'> <source> <address domain='0x0' bus='0x1' slot='0x00' function='0x0'/> </source> </hostdev> When I execute virsh attach-device testvm hga.xml I get the following error: Fehler: Fehler beim Anhängen des Geräts von hga.xml (error: error while attaching device from hga.xml) Fehler: Interner Fehler: Kann Qemu-Befehl 'device_add' nicht ausführen: Device initialization failed. (error: internal error: cannot execute 'device_add': In the syslog of the host I get the following errors: Oct 7 11:32:34 artemis2 journal: Interner Fehler: Kann Qemu-Befehl 'device_add' nicht ausführen: Device initialization failed. Oct 7 11:32:53 artemis2 kernel: ffff880220fadd10 ffffffff810f924a ffff880220fadd28 0000000000000246 Oct 7 11:32:53 artemis2 kernel: ffff8800bd07a400 000000007265756d ffffffffa0866ee0 0000000000000000 Oct 7 11:32:53 artemis2 kernel: Call Trace: Oct 7 11:32:53 artemis2 kernel: [<ffffffff815e237b>] dump_stack+0x19/0x1b Oct 7 11:32:53 artemis2 kernel: [<ffffffff810f924a>] __setup_irq+0x54a/0x570 Oct 7 11:32:53 artemis2 kernel: [<ffffffffa0866ee0>] ? virqfd_wakeup+0xc0/0xc0 [vfio_pci] Oct 7 11:32:53 artemis2 kernel: [<ffffffff810f93cc>] request_threaded_irq+0xcc/0x170 Oct 7 11:32:53 artemis2 kernel: [<ffffffffa0866b8a>] vfio_intx_set_signal+0x10a/0x1f0 [vfio_pci] Oct 7 11:32:53 artemis2 kernel: [<ffffffffa08670e8>] vfio_pci_set_intx_trigger+0xc8/0x1b0 [vfio_pci] Oct 7 11:32:53 artemis2 kernel: [<ffffffffa086799b>] vfio_pci_set_irqs_ioctl+0x7b/0x100 [vfio_pci] Oct 7 11:32:53 artemis2 kernel: [<ffffffff8115d352>] ? memdup_user+0x42/0x70 Oct 7 11:32:53 artemis2 kernel: [<ffffffffa0865e74>] vfio_pci_ioctl+0x314/0xa00 [vfio_pci] Oct 7 11:32:53 artemis2 kernel: [<ffffffff81194b6a>] ? kmem_cache_alloc+0x1ba/0x1d0 Oct 7 11:32:53 artemis2 kernel: [<ffffffff811b156c>] ? get_empty_filp+0x5c/0x190 Oct 7 11:32:53 artemis2 kernel: [<ffffffff811f1891>] ? fsnotify+0x251/0x350 Oct 7 11:32:53 artemis2 kernel: [<ffffffff81251806>] ? security_file_alloc+0x16/0x20 Oct 7 11:32:53 artemis2 kernel: [<ffffffff811b15a0>] ? get_empty_filp+0x90/0x190 Oct 7 11:32:53 artemis2 kernel: [<ffffffffa084f163>] vfio_device_fops_unl_ioctl+0x23/0x30 [vfio] Oct 7 11:32:53 artemis2 kernel: [<ffffffff811c2af5>] do_vfs_ioctl+0x2e5/0x4c0 Oct 7 11:32:53 artemis2 kernel: [<ffffffff811af780>] ? vfs_write+0x160/0x1e0 Oct 7 11:32:53 artemis2 kernel: [<ffffffff811c2d71>] SyS_ioctl+0xa1/0xc0 Oct 7 11:32:53 artemis2 kernel: [<ffffffff815f2a59>] system_call_fastpath+0x16/0x1b Oct 7 11:32:53 artemis2 journal: Interner Fehler: Kann Qemu-Befehl 'device_add' nicht ausführen: Device initialization failed. My version-information: OS: CentOS Linux release 7.0.1406 (Core) Kernel: Linux artemis2 3.10.0-123.8.1.el7.x86_64 libvirt-daemon-driver-secret-1.2.8-1.el7.centos.x86_64 libvirt-1.2.8-1.el7.centos.x86_64 libvirt-daemon-kvm-1.2.8-1.el7.centos.x86_64 libvirt-client-1.2.8-1.el7.centos.x86_64 libvirt-daemon-driver-qemu-1.2.8-1.el7.centos.x86_64 libvirt-daemon-config-network-1.2.8-1.el7.centos.x86_64 libvirt-daemon-driver-interface-1.2.8-1.el7.centos.x86_64 libvirt-daemon-1.2.8-1.el7.centos.x86_64 libvirt-daemon-driver-nodedev-1.2.8-1.el7.centos.x86_64 libvirt-daemon-driver-storage-1.2.8-1.el7.centos.x86_64 libvirt-daemon-config-nwfilter-1.2.8-1.el7.centos.x86_64 libvirt-daemon-driver-network-1.2.8-1.el7.centos.x86_64 libvirt-daemon-driver-lxc-1.2.8-1.el7.centos.x86_64 libvirt-daemon-driver-nwfilter-1.2.8-1.el7.centos.x86_64 The PCI-Card I want to passthrough: 02:00.0 Ethernet controller: PLX Technology, Inc. Device 235e Subsystem: PLX Technology, Inc. Device 235e Control: I/O- Mem- BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx- Status: Cap+ 66MHz- UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx- Interrupt: pin A routed to IRQ 16 Region 0: Memory at f0810000 (32-bit, non-prefetchable) [disabled] [size=256] Region 1: I/O ports at e000 [disabled] [size=256] Region 2: Memory at e0000000 (32-bit, non-prefetchable) [disabled] [size=256M] Region 3: Memory at f0000000 (32-bit, non-prefetchable) [disabled] [size=8M] Expansion ROM at f0800000 [disabled] [size=64K] Capabilities: [40] Power Management version 1 Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot-,D3cold-) Status: D0 NoSoftRst- PME-Enable- DSel=0 DScale=0 PME- Capabilities: [48] CompactPCI hot-swap <?> Capabilities: [4c] Vital Product Data Unknown large resource type 35, will not decode more. Kernel driver in use: vfio-pci Please help. Best regards Michael Weis PROPRIETARY: This e-mail contains proprietary information some or all of which may be legally privileged. It is intended for the recipient only. If an addressing or transmission error has misdirected this e-mail, please notify the authority by replying to this e-mail. If you are not the intended recipient you must not use, disclose, distribute, copy, print, or rely on this e-mail.

1 0

[libvirt-users] FreeBSD 10-RELEASE-amd64(on arch x86_64)
by Denis Menshikov 06 Oct '14

06 Oct '14

Good day, would like to install the latest version of libvirt collected from sources for use bhyve. Download ftp://libvirt.org/libvirt/libvirt-1.2.9.tar.gz, system... uname -a FreeBSD vmserver 10.0-RELEASE FreeBSD 10.0-RELEASE #0 r260789: Thu Jan 16 22:34:59 UTC 2014 root@snap.freebsd.org:/usr/obj/usr/src/sys/GENERIC amd64 my actions: ./configure make sudo make install next... make all-recursive Making all in . Making all in gnulib/lib GEN alloca.h GEN c++defs.h GEN warn-on-use.h GEN arg-nonnull.h GEN arpa/inet.h GEN byteswap.h GEN configmake.h GEN fcntl.h GEN langinfo.h GEN locale.h GEN netdb.h GEN poll.h GEN pty.h GEN signal.h GEN unused-parameter.h GEN stdio.h GEN stdlib.h GEN string.h GEN strings.h GEN sys/ioctl.h GEN sys/select.h GEN sys/socket.h GEN sys/stat.h GEN sys/time.h GEN sys/types.h GEN sys/uio.h GEN sys/utsname.h GEN sys/wait.h GEN termios.h GEN time.h GEN unistd.h GEN wchar.h GEN wctype.h make all-am CC allocator.lo CC areadlink.lo CC base64.lo CC binary-io.lo CC bitrotate.lo CC c-ctype.lo CC c-strcasecmp.lo CC c-strncasecmp.lo CC c-strcasestr.lo CC careadlinkat.lo CC cloexec.lo CC count-one-bits.lo CC md5.lo CC sha256.lo CC dirname-lgpl.lo CC basename-lgpl.lo CC stripslash.lo CC fd-hook.lo CC freading.lo CC getugroups.lo CC localcharset.lo CC glthread/lock.lo CC malloca.lo CC mgetgroups.lo CC nonblocking.lo CC passfd.lo CC physmem.lo CC pipe2.lo CC sig-handler.lo CC sockets.lo CC stat-time.lo CC strnlen1.lo CC sys_socket.lo CC tempname.lo CC glthread/threadlib.lo CC unistd.lo CC wctype-h.lo CC xsize.lo CC asnprintf.lo CC canonicalize-lgpl.lo CC fclose.lo CC fdatasync.lo CC fflush.lo CC fseek.lo CC fseeko.lo CC getgroups.lo CC ioctl.lo CC mktime.lo CC openpty.lo CC printf-args.lo CC printf-parse.lo CC pthread_sigmask.lo CC random_r.lo CC rawmemchr.lo CC regex.lo In file included from regex.c:70: ./regex_internal.c:1392:11: warning: comparison of unsigned expression < 0 is always false [-Wtautological-compare] if (idx < 0 || idx >= set->nelem) ~~~ ^ ~ 1 warning generated. CC secure_getenv.lo CC timegm.lo CC vasnprintf.lo CCLD libgnu.la GEN charset.alias GEN ref-add.sed GEN ref-del.sed Making all in include Making all in libvirt Making all in src GEN libvirt_access.xml GEN libvirt_access_qemu.xml GEN libvirt_access_lxc.xml GEN libvirt_access.syms GEN libvirt_access_qemu.syms GEN libvirt_access_lxc.syms GEN libvirt.syms GEN libvirt.def Next, the compile stop, activity on the CPU no. What am I doing wrong? -- Denis Menshikov

2 1